A Review of Inter-Modular Connections for Volumetric Cross-Laminated Timber Modular Buildings

Abstract

The application of volumetric modular construction using Cross-Laminated Timber (CLT) has emerged as a sustainable and efficient alternative to traditional building methods, especially in residential and mid-rise structures. However, the widespread adoption of this technology remains limited due to the lack of standardized inter-modular connection systems. This paper presents a comprehensive state-of-the-art review of inter-modular connections used in volumetric CLT modular buildings. This review aims to evaluate the inter-modular connections by developing performance objectives and identifying gaps in knowledge of volumetric CLT inter-modular connections. It begins with an overview of global CLT modular construction trends, highlighting geographic distribution, structural demands, and environmental hazards such as seismic and wind exposure. Seven representative connection systems were identified from the literature and assessed using a multi-criteria framework comprising structural performance, manufacturing feasibility, on-site construction efficiency, and experimental and numerical evaluation. Each connection was scored according to defined evaluation metrics, and the results were provided to identify key strengths and limitations. The top-performing systems demonstrated superior resilience, modular adaptability, and validation through experimental testing and simulation. The paper identified critical research gaps, including limited performance data available for seismic applications, challenges in disassembly and reuse specifications, and the need for adaptable, damage-tolerant systems to enhance building structural performance. These findings provide a reference evaluation methodology for future development of inter-modular connections, to expand the applicability of volumetric CLT modular construction in moderate and high seismic and wind hazard regions.

1. Introduction

The Architecture, Engineering, and Construction (AEC) industry has multiple efforts underway to improve material efficiency, reduce costs, and reduce greenhouse gas emissions, all while meeting the large housing demands that exist in our urban cores today. Modular construction has shown promising results to meet these performance objectives. The prefabrication process and the capacity to reutilize materials associated with modular construction methods resulted in lower lifecycle emissions of modular construction buildings compared to those of traditional construction [1]. Furthermore, modular construction takes advantage of recent technological development and automation processes within the AEC, which can further support the rapid expansion of the modular construction industry [2].

Modular construction is a broad methodology that is equally applied to multiple accelerated construction concepts, especially for timber structures [3]. The most common definitions describe modular construction as a process where the elements are built off-site, under controlled conditions, and subjected to higher quality control. The products of the production process are transported and assembled at a final location under the exact same specifications as were designed [4,5,6].

Modular buildings can be assembled using different types of elements, linear, planar, or volumetric units [7,8]. For the case of volumetric modular construction, the modular units include internal finishes, ranging between 50% and 90% of the total work completed during the prefabrication process off-site [9]. Independent of the material, but based on the level of prefabrication, some cases of timber modular construction reported up to 95% of completion [10,11,12]. This off-site construction reduces the number of people required on-site, the time required for the building to be occupiable, and delays due to design modifications [6]. All of these benefits can reduce the timeline of projects by 50% [13].

Despite these massive benefits, major gaps in knowledge persist for timber volumetric modular construction. Namely, the need for further testing and analysis of modular building performance, promoting the development of standardized solutions to enhance versatility in designing various floor plan configurations [11]. Specifically, there is a lack of solutions for inter-modular connections for timber modular buildings where units are stacked together and anchored to an exoskeleton or an internal core. These structural solutions require connections with complex demand requirements that have not yet been fully developed.

In moderate and high seismic regions, inter-modular connections represent a critical component for modular buildings. These connections are responsible for transferring loads between the modular units, enabling diaphragm action within the building, and transferring the horizontal loads to the lateral force resisting system. The intra-modular connections enable the assembly of the building and should be designed to enable the components of the modular units to remain in the elastic range or avoid internal damage.

Previous research on steel modular buildings [14,15] exhibited the importance of inter-modular connections in the global response of modular buildings. In the timber industry, a small number of connections were studied, and there is a need for future research on connection systems, including experimental testing and numerical simulations for timber modular buildings.

The timber construction industry has shown interest in developing new methodologies for applying volumetric CLT modular units at various building scales. As a complementary action, technical guidelines have been developed to support the design and evaluation of modular buildings [16,17]. However, the published documentation does not cover the specific structural and connection systems requirements of the modular buildings. This gap in the guidelines has been identified by other researchers as well [18,19] who mention that connections for the CLT modules require additional assessment, and the lack of regulations and design specifications for modular buildings.

Considering the lack of standardized details and design methodologies to support the analysis and design of CLT volumetric modular structures, this paper presents a state-of-the-art review of the existing literature on the published inter-modular connections for volumetric CLT modular buildings. A description of the connections and the experimental and numerical tests performed are presented to help researchers and engineers understand the performance requirements of CLT inter-modular connections. In addition, the authors developed a methodology to evaluate the published inter-modular connections across four key attributes critical for the success of modular buildings: structural, manufacturing, construction, and experimental. The objectives of this research project are to: (1) review and evaluate inter-modular connections developed for volumetric CLT modular building construction; (2) develop performance objectives of inter-modular connections to be used in design of modular buildings; and (3) identify gaps in knowledge of volumetric CLT inter-modular connections that are research priorities to enable modular building design in high seismic regions.

CLT is a structural building material that was developed in the 1980s in Europe. Since then, extensive research has been conducted to enhance its performance. In the 2010s, CLT was used for tall timber buildings with the approval of standards and guidelines developed to satisfy regulatory requirements [20,21,22,23]. Over the last decade, research on CLT has continued in order to gain more knowledge on the fire resistance, acoustics, and sustainability behavior of the material [24]. The use of volumetric CLT modular units has increased worldwide [25,26,27], driven by the growing demand for housing options.

2. Methodology

The authors performed a state-of-the-art review on inter-modular connections for volumetric CLT modular buildings to identify, evaluate, and synthesize the published literature. The applied methodology was founded on previous approaches used to review timber structures and inter-modular connections [11,14,28,29,30]. The review involved two major steps: a literature review with data extraction, and an evaluation of the inter-modular connections’ performance.

2.1. Literature Review and Data Collection

To define the scope of the review process, no specific geographic boundaries were defined for the literature selection. Given the nature of the emerging construction technology, selected references were chosen after 2005, following the findings of [29], which described 2005 as the beginning of an increased utilization of CLT as a prefabricated component in medium- and high-rise buildings. The source selection was performed through a deep search within different sources, including academic databases, published journal articles, conference papers, technical reports, and proprietary design guidelines. The strings used to perform the search were “CLT modular construction connections”, “CLT inter-modular connections”, “CLT intermodular connections”, “CLT volumetric connection”, “volumetric CLT buildings”, “volumetric CLT connectors”, “CLT module connection”, “CLT modular unit”, “CLT modular unit connections”, “volumetric prefabricated modules timber”, “timber prefabricated buildings”, “timber prefabricated buildings CLT”, “connection prefabricated CLT module”, “mass timber inter-module connections”, “volumetric CLT modular building”, “CLT volumetric modules”, “industrialized construction”, “industrialized construction CLT connections”, “volumetric industrialized construction CLT connections”, “volumetric industrialized construction CLT inter-modular connections”, “off-site construction CLT connections”, “off-site construction CLT inter-modular”, “off-site construction CLT intermodule”, “off-site construction CLT intermodular”, and a combination of these words.

The keywords were searched in Google Scholar, Web of Science, and ScienceDirect databases. Additionally, a complementary data search was conducted on professional social media platforms (LinkedIn), online journals, and industry websites. In addition, email requests for data information were sent to architectural firms, structural designers, and modular CLT manufacturers. The resources were mainly in English; however, when resources were in another language, the automatic translation from Google Chrome was used. Only sources that included all parameters and were published were included in the evaluation.

Once the data was collected, the screening process was performed based on the relevance of the resources to volumetric CLT modular buildings. The collected data was classified into information about the published volumetric CLT modular buildings and information related to the details of the inter-modular connections for the description and performance evaluation. With the filtered and classified data, a critical and technical analysis was performed to identify the components and evaluate the performance of the inter-modular connections. A synthesis of the information was provided to highlight the behavior of the connection, primary features, and the main outcomes from published experimental and numerical testing of each connection.

Seismic and Wind Hazard Definition

Based on the geographic location of the modular buildings, the seismic hazard was estimated by obtaining the peak ground acceleration (PGA) from the Global Earthquake Model provided by the Global Seismic Hazard Map [31]. The seismic hazard was classified into low, moderate, and high categories [32]. The regions with low seismic ground motion hazard were associated with places with a PGA less than or equal to 0.1 g. Regions with moderate seismic ground motion hazard were classified between 0.1 g and 0.2 g. Finally, regions with high seismic ground motion hazard were selected for places with PGA equal to or greater than 0.2 g. The PGA used for estimating the seismic hazard has a 2% probability of being exceeded in 50 years.

Similarly, the wind hazard was estimated by obtaining the maximum wind speed at 10 m height from the New European Wind Atlas project [33] and the wind speed for a risk category II from the ASCE Hazard Tool [34]. The wind hazard under normal conditions [35] was classified in

Table 1. Wind speed ranges to define wind hazard.

Case	Range	Wind Speed (m/s)
Normal conditions	Non-threatening	minimum
	Very low	<9
	Low	9–11
	Moderate	11–17.5
	High	17.5–25.5
	Extreme	>25.5
Hurricane	Very dangerous	33–42.5
	Extremely dangerous	42.5–49
	Devastating damage	49–57.5
	Catastrophic damage	57.5–70
	Total damage	>70

, and additional ranges were included to provide the perspective of the wind speeds that could be reached during hurricane conditions [36]. The wind speed from the ASCE Hazard Tool [34] corresponds to a three-second gust wind speed at 10 m above the ground in exposure C. The wind speed corresponds to approximately a 7% exceedance probability in 50 years.

2.2. Evaluation of Inter-Modular Connections for CLT Modular Buildings

An evaluation process for the published inter-modular connections for volumetric CLT modular buildings was proposed following the methodologies developed by [14,28]. The evaluation was performed to compare the features of the connection systems. With the proposed methodology, the connection properties can be highlighted to identify the advantages or disadvantages of each inter-modular connection. Additionally, the evaluation provides a foundation for future designs, identifying areas that could be improved. Each connection was designed or tested under different circumstances to fulfill a similar objective; however, the performance demands for the connections were not the same. Based on this principle, a quantitative comparison is not appropriate; therefore, a qualitative evaluation was conducted for a comparison between the connections.

Four categories were defined to evaluate several aspects of the connection and its conceptual design (

Table 2. Evaluation metrics and scoring system for the performance assessment of inter-modular connections.

Structural Metrics
Metric	Score	Description
VC	0	Does not provide vertical connectivity.
	1	Provides vertical connectivity to prevent module uplift.
HC	0	Does not provide horizontal connectivity.
	1	Provides horizontal connectivity with axial and shear capacity.
CR	0	Provides low or non-energy dissipation capacity, and the damage is localized in the timber elements.
	1	Provides a moderate energy dissipation capacity, and an intermediate or low level of damage is identified in the surrounding timber elements.
	2	Provides good energy dissipation capacity, and limited or no damage is produced to surrounding timber elements. The connection can be replaced.
DF	0	Limited design flexibility. Designed to be positioned at a specific location within the modular unit.
	1	Moderate design flexibility. Requires modifications and can be adapted for use in any location within the building.
	2	Good design flexibility. The connection is standard for any location within the building.
Manufacturing Metrics
Metric	Score	Description
CC	0	Requires the manufacture of a large number of parts with a complex geometry or a complex manufacturing process. The mass production of components might be time-consuming.
	1	Moderate number of parts, with a regular geometry that does not require an intensive manufacturing process. The mass production of components requires moderate time operations.
	2	A small number of parts with simple and regular geometry without complex manufacturing processes. Mass production is simple, with minimal time-consuming operations.
CI	0	Complex manufacturing processes required to integrate the connection components (e.g., welding).
	1	Intermediate difficulty for integration of connection components (e.g., drilling, welding of small parts).
	2	Simple integration of connection components (e.g., fastening, labeling, aligning elements).
IP	0	Rigorous installation process of the inter-modular connection in the modular units (e.g., welding components into embedded parts, fitting heavy components).
	1	Intermediate difficulty for the installation of the inter-modular connection (e.g., drilling timber panels, fastening or screwing of components, laser cutting of panels).
	2	Simple installation of the inter-modular connections (e.g., fastening single elements, aligning components).
Construction Metrics
Metric	Score	Description
DA	0	Complex construction methods, no self-aligning/self-locating features, a large number of tasks, difficult access, and complex tooling.
	1	Intermediate difficulty for construction methods, self-aligning/self-locating features, moderate number of tasks, moderate access, and moderate tooling.
	2	Simple construction methods, efficient self-aligning/self-locating features, a small number of tasks, easy access, and simple tooling.
DD	0	Difficult process required to disassemble, limited access. Requires additional work before reusing.
	1	Intermediate difficulty to disassemble, a few parts of the modular unit may need to be replaced before reuse, with minimum access limitations.
	2	Easy to disassemble, the modular units can be immediately reused or require small adjustments, adequate access.
TC	0	Limited tolerance control, the construction process requires corrections to fit the modules.
	1	Intermediate tolerance control, the construction process requires minor adjustments or modular alignment to fit the modular units.
	2	Adequate tolerance control, the modular connections effectively handle the required adjustments during the construction process.
Experimental and Numerical Metrics
Metric	Score	Description
CET	0	No experimental tests have been performed on any component of the inter-modular connection.
	1	Monotonic or cyclic experimental tests have been performed over the components of the inter-modular connection.
FSET	0	No experimental test at full scale has been performed with the inter-modular connection.
	1	Monotonic or cyclic tests at full scale have been performed with the inter-modular connection.
NME	0	The connection has not been numerically simulated.
	1	An elastic numerical model of the inter-modular connection has been developed to simulate its behavior.
	2	A nonlinear numerical model of the inter-modular connection has been developed and validated against experimental test results.
AMD	0	No analytical model has been developed for the inter-modular connection.
	1	An analytical model has been developed to simplify the design of the modular building.

). The authors expanded upon the criterion provided by [14,28], to evaluate the data from the literature. Each feature is described in Table 2 with the corresponding score system. The scoring system considers a grading methodology within a range from zero to one depending on whether the given conditions were met, and a second scale from zero to two to describe the qualitative properties of the connection in terms of quality, efficiency, or advantages. The maximum possible score for a connection system equals 23, while the minimum possible score equals 0. Additional explanation about each performance metric is provided in the following Section 2.2.1, Section 2.2.2, Section 2.2.3 and Section 2.2.4. The authors acknowledge that the scores assigned to each connection system were determined through a subjective process based on the published data of the connections and served as a reference value based on the provided interpretation. Other interpretations of the qualitative parameters might provide slight differences in the connection evaluation.

2.2.1. Structural Evaluation

The structural performance of the inter-modular connections is fundamental to defining the diaphragm behavior and the global structural response of the modular buildings [37,38]. The inter-modular connections were evaluated using four parameters, which consider the connections’ performance, design, and structural concept. To evaluate the performance, the vertical connection (VC), and horizontal connection (HC) were evaluated as individual components to consider an independent response for each direction. Due to seismic and wind demands, the connection should be capable of resisting uplift in the vertical direction. At the same time, the connection should have the capability to resist axial and shear forces in the horizontal direction to develop diaphragm action within the module.

In addition, the connection resilience (CR) was evaluated based on the energy dissipation capacity and the ability to concentrate the damage in the connection to reduce the damage in the timber elements in case of large seismic or wind demands. The last parameter, design flexibility (DF), evaluated the conceptual design of the connection by considering the feasibility of standardizing the connection to facilitate modifications in the modular buildings’ configurations. It evaluated the limitations of the connection related to the location of the connection in the modular unit and the module’s location in the building plan view (corner, border, or central unit).

2.2.2. Manufacturing Evaluation

The manufacturing process for modular construction in off-site industries is the fundamental procedure to improve efficiency. The effective machine operation and adequate sequence of operations are continuously monitored to control and guarantee the best performance [39]. To evaluate the manufacturing process, three parameters were considered. The connection complexity (CC) was based on the number of operations or the difficulty of fabricating the connection components to assess the feasibility of large-scale production of the connection. Components integration (CI) evaluated the assembly process of the connections, including any mechanical process required to assemble the connections. The installation procedure (IP) evaluated the complementary drilling, welding, or any additional machining procedure required for the installation of the connections in modular units. All these procedures were developed off-site, in an industrial environment.

2.2.3. Construction Evaluation

The construction process was evaluated through the design concept of the connection to facilitate on-site operations. The design for assembly (DfA) and design for disassembly (DfD) represented an important quality for modular construction, which could benefit the reusability of the structures with the development of new connections to allow such operations [30]. The DfA evaluated the amount of work that should be completed on-site. This parameter evaluated if the inter-modular connection provided simple methods to assemble the modular units and facilitated a rapid installation to reduce the time required for on-site construction. Complementary, DfD assessed the feasibility of deconstructing the modular building and the accessibility for performing such tasks. It also evaluated the capacity of reusing the modular units without extensive additional work.

In addition, the construction tolerances of modular buildings should be considered during the design stage, defining allowable limits due to the effects of imperfections, and how they should be included during the analysis of the modular buildings [40]. The tolerance control (TC) evaluated if the connection system allowed for a certain tolerance level when fitting the modular units during construction. It considers whether the connection provides sufficient tolerance to accommodate the modular units and prevent the propagation of geometric imperfections to adjacent modular units.

2.2.4. Experimental and Numerical Evaluation

From the reviewed data, the experimental and numerical evaluation of the inter-modular connections provided additional support for the connection design. The published data on these connections within these categories demonstrated systematic and scientific evaluation of the connection behavior that can be used for structural design. The component experimental testing (CET) evaluated whether monotonic or cyclic experimental tests were performed at the component level to validate the expected behavior of the individual components of the inter-modular connection. This includes material strength verifications, fastener testing, or any other individual component to verify its performance. The full-scale experimental testing (FSET) evaluated whether the inter-modular connection was installed and tested in full-scale under monotonic or cyclic loading conditions to assess its structural performance.

Additionally, the numerical model for evaluation (NME) was defined as an additional parameter to determine if a numerical model of the inter-modular connection had been developed to simulate its behavior. Special attention was given to inter-modular connections when nonlinear models were developed to benchmark the experimental results. In addition, the analytical model for design (AMD) was considered as a parameter to evaluate whether an analytical model of the connection had been developed to simplify the analysis and design of modular buildings with inter-modular connections.

This methodology combined different approaches to provide a comprehensive summary of the published data, offering a broader perspective on the research gaps that require attention to support the incorporation of volumetric CLT modular buildings in the construction markets of regions with high wind and seismic hazards.

3. Results

From the published literature and online resources, a set of volumetric CLT modular buildings was identified and listed in

Table 3. Volumetric CLT modular units reported in the literature.

Building Name	City, Country	Completion Year	# Stories	# Modules	Seismic Hazard	Wind Hazard	Reference
BMW Alpenhotel Ammerwald	Ammerwald, Austria	2009	5 (concrete first two stories)	96	Moderate	High	[7,41,42]
Senior Citizens’ Home in Hallein	Hallein, Austria	2013	5 (concrete ground floor)	136	Moderate	High	[41]
Student Hostel in Heidelberg	Heidelberg, Germany	2013	5	265	High	Moderate	[7,41]
Moxy Hotels	Multiple countries	2014	7	200	-	-	[7]
Adoma Apartments	Toulouse, France	2015	4	56	Low	High	[7,41]
Frankfurt European School	Frankfurt Am Main, Germany	2015	3	98	High	Moderate	[7,41,43]
Puuokoka House 1	Vainonkatu, Finland	2015	8	116	Low	Moderate	[7,44]
Puuokoka House 2	Vainonkatu, Finland	2015	7	91	Low	Moderate	[7,44]
Puuokoka House 3	Vainonkatu, Finland	2015	6	71	Low	Moderate	[7,44]
Steigart Strasse Refugee Settlement	Hannover, Germany	2015	2	185	Low	High	[7,41]
Treet	Bergen, Norway	2015	14	62	Moderate	High	[45,46,47]
Integrated Comprehensive School	Frankfurt am Main, Germany	2016	3	90	High	Moderate	[42,43]
Wohnen 500	Mäder, Austria	2016	3	60	High	High	[41]
Hotel Katharinenhof	Dornbirn, Austria	2017	4 (concrete ground floor)	39	High	High	[41]
Hotel Revier	Lenzerheide, Switzerland	2017	4 (concrete ground floor)	96	High	High	[42]
Woodie Student Hostel	Hamburg, Germany	2017	7 (concrete ground floor)	371	Low	High	[7,41,42]
Hotel Jakarta	Amsterdam, Netherlands	2018	9	176	Low	Moderate	[7]
DAS Kelo	Rovaniemi, Finland	2019	8	-	Low	Moderate	[48,49]
Daycare Center	Frankfurt am Main, Germany	2019	2	50	High	Moderate	[42,43]
Hotel Bergamo	Ludwigsburg, Germany	2019	4	440	Moderate	High	[42]
Josefhof Health Centre	Graz, Austria	2019	2	120	High	High	[42]
Toimela	Nurmijärvi, Finland	2019	4	-	Low	High	[48,49]
Kirkkonummen Konsulintorni	Kirkkonummi, Finland	2020	4	-	Low	High	[48,49]
Lutterterrasse Student Residence	Göttingen, Germany	2020	5 (concrete ground floor)	265	Low	High	[42]
Mannisenrinteen Puumanni, Building A	Jyväskylä, Finland	2020	4	-	Low	Moderate	[48,49]
Mannisenrinteen Puumanni, Building B	Jyväskylä, Finland	2020	4	-	Low	Moderate	[48,49]
Office Kaufmann Building Systems	Reuthe, Austria	2020	3 (concrete ground floor)	32	High	High	[42,43]
Watts Grove	Tower Hamlets, England	2020	6	-	Low	High	[50,51]
Kaarna	Kuopio, Finland	2021	7	-	Low	High	[49]
Luisenblock West	Berlin, Germany	2021	7	460	Low	High	[42]
Rautalepänkatu 2 Building A	Tampere, Finland	2021	4	-	Low	High	[49]
Rautalepänkatu 2 Building B	Tampere, Finland	2021	4	-	Low	High	[49]
Tampereen Härmälänsydän	Tampere, Finland	2021	4	-	Low	High	[48,49]
Tampereen Kaupin puukerrostalo	Tampere, Finland	2021	8	-	Low	High	[48,49]
Vaasan Viherlehto	Vaasa, Finland	2021	6	-	Low	High	[48,49]
Koelmalaan	Alkmaar, Netherlands	2022	5	260	Low	High	[52,53]
Lumipuu, Building A	Tampere, Finland	2022	6	-	Low	High	[48,49]
Lumipuu, Building B	Tampere, Finland	2022	6	-	Low	High	[48,49]
Nila	Kuopio, Finland	2022	7	-	Low	High	[49]
Pyssysepänkaari 3	Kirkkonummi, Finland	2022	5	-	Low	High	[49]
Boarding School Wood Technology Centre	Kuchl, Austria	2023	7 (concrete ground floor)	82	Moderate	High	[42]
Vogewosi, Flurgasse—Housing 500	Feldkirch, Austria	2023	3	-	High	High	[42]
Campus Valkenvoortweg Waalwijk	Waalwijk, Netherlands	2024	3	-	High	High	[52,54]
Housing in Big Sky	Big Sky, United States	2025	3	120	High	Extreme	[43]

to evaluate the geographic distribution of the buildings, the year of completion, the number of stories, and the number of modules used to assemble the buildings. Based on the location of the buildings, the PGA and the wind speed at 10 m above the ground in exposure C were obtained from the hazard models [31,33] to identify the seismic and wind hazard for each modular building. The list focuses exclusively on pure volumetric CLT modular buildings, excluding cases where other types of modular systems were used.

From the identified volumetric CLT modular buildings in Table 3, 60% of the cases are in low-seismic regions, 12% of the cases are in moderate seismic regions, and 28% of the cases are in high-seismic regions. In terms of the wind hazard, 26% of the buildings are in a moderate wind hazard region, 72% in a high wind hazard region, and 2% in an extreme wind hazard region. A visual reference of the location of the modular buildings and their seismic hazard is provided in Figure 1a,b. A similar reference for the wind hazard distribution is provided in Figure 2. Most of the CLT volumetric units reported in the literature are distributed across different countries in Europe, especially those with lower seismic and wind demands. According to the reported information in the literature, many of these buildings were intended for residential use. These results demonstrate the importance of having connection details that can resist seismic demands.

Table 4. Inter-modular connections for volumetric modular buildings.

ID	Connection Description	Figure	Fasteners Detail	Ref.
IMC-1	The connection system uses steel plates with screws to fasten adjacent modules. A soundproofing mechanism was proposed by installing elastic bearings between modular units that lack internal connectivity and have no soundproofing for modules within the same occupational unit. Steel plates were proposed for horizontal and vertical inter-modular connections to withstand shear forces, while steel rods were suggested to control uplift forces within the modules.		Fastener size between 6 and 12 mm screws	[55]
IMC-2	An interlocking connection system was proposed for modular volumetric construction. It consists of steel 3D-printed tensile and shear connections, composed of male and female components that self-lock, restraining movement in their primary and secondary working directions. The construction process requires a specific order to slide and stack the modules.		Screws used for the connections HBSP12120 and LBS7100	[56,57,58,59]
IMC-3	A connection system named X-RAD, designed for CLT structures that can be applied to both 2D and 3D modular construction. The X-RAD system was designed as a point-to-point mechanical connection located at the corners of the CLT panels. It is composed of an outer metallic envelope, an internal steel plate, an inner core made of hardwood, a pair of horizontal bolts, and a set of six self-tapping screws.		M12, M16 bolts, and VGS 11 × 350 screws	[60,61,62,63,64,65,66]
IMC-4	The inter-modular connection proposed consists of a steel T-shaped angle plate with pins. The steel pins were designed to fit into steel cones that should be cast into the prefabricated floor slab. For the assembly of the modular building, the steel plates are screwed on top of two adjacent side walls prior to stacking the modules of the next story. Once the first module of the next story is placed, the T-shape is secured on one side of the slab of that module.		Screws *	[67]
IMC-5	The proposed dowel-type connection is designed to join the floor panel of an upper CLT volumetric module to the ceiling panel of the module located directly beneath it, enabling effective shear transfer between vertically adjacent modules in a stacked modular building system.		25.4 mm Read Oak and Yellow Birch dowels, and 8 mm fully threaded screws (SDCF22614 and SDCF22858)	[68]
IMC-6	Two independent connections were developed as the inter-modular connection for volumetric stackable timber modules. The vertical connection was attached to the walls of the upper and lower modules to connect them through a bolt, preventing the uplift and horizontal sliding of the modules. The horizontal connection was designed as a damper device using low-yield steel to dissipate energy through shear deformation.		8 × 160 mm screws	[69]
IMC-7	Straviwood ModuLink consists of two L-shaped steel brackets, one of which is shorter than the other. They are linked by two bolts M8 × 65 and contain two elastomeric bearing (foam) blocks to provide acoustic-isolation capacity. A locknut secures the plates in place and ensures adequate pre-compression of the foam block. It was developed as a horizontal connection for contiguous modular CLT constructions. Patent number BE 1030721.		3 HECO-TOPIX-plus 10 × 100, flange head screws per side	[70,71,72,73]

* No detail was provided about the fasteners used for the connection.

provides a summary of developed inter-modular connections from literature. These connections were selected based on their capacity to transfer vertical and lateral loads between volumetric modules, having testing data on them under seismic demands, and reported volumetric modular buildings where the connections have been used for their construction. From the identified inter-modular connections, a short description of the connection system was provided, with a schematic figure of the connections to provide a better understanding of the system, and the details of the types of fasteners used for the connection with the CLT modules. The filters applied resulted in seven connections identified from the literature. A unique identifier was assigned to each connection, which was used throughout the analysis.

3.1. Structural Evaluation

The structural performance of volumetric CLT modular buildings is directly related to the design and response of the inter-modular connections when buildings are subjected to lateral loads from seismic or wind demands [37,38,40]. These components serve as critical load path mechanisms for both vertical and horizontal forces, and their behavior directly affects the global stability of modular structures.

Table 5. Individual scores assigned for structural evaluation. Structural metrics abbreviations from Table 2; Connection ID from Table 4.

	Conn. ID	IMC-1	IMC-2	IMC-3	IMC-4	IMC-5	IMC-6	IMC-7
Structural metrics	VC	1	1	1	1	0	1	0
	HC	1	1	1	1	1	1	1
	CR	0	2	1	0	1	1	0
	DF	1	2	2	0	1	1	2
	Total	3	6	5	2	3	4	3

provides the individual scores assigned to each connection, while Figure 3 illustrates the cumulative scores.

IMC-2 obtained the highest structural score due to its well-integrated design, which addresses all structural criteria. This connection was designed with an interlocking mechanism to resist tensile and shear stresses, thereby restricting movement in both its primary and secondary directions. Based on the concept of strong fasteners and weak metal connections, it was designed to absorb damage within the steel connection components, thus protecting the CLT modular units (CR = 2). Additionally, its design flexibility (DF = 2) ensures its applicability across different modular layouts, making it a robust and adaptable solution for multistory buildings with variable configurations.

IMC-3 is unique in its corner-based configuration and capacity to simultaneously resist shear and uplift due to its 3D screw layout. Its intermediate resilience score (CR = 1) reflects the presence of energy dissipation and controlled damage mechanisms, which, unfortunately, include damage to the timber panels. As a highly standardized solution, it offers design flexibility (DF = 2) and has already been implemented in several experimental and prototypical CLT modular structures [64].

IMC-6 offers both vertical and horizontal connectivity, with intermediate CR and DF. A low-yield steel (LYS) damper was used for energy dissipation in the horizontal component of the connection. This damper was designed to yield under seismic loads and can be replaced in case of damage, providing resilient features to the modular building (CR = 1). The design flexibility was reduced (DF = 1), indicating limitations in modular building configuration and standardization of connections.

IMC-1 was conceptualized using steel plates for vertical and horizontal connections and continuous tie-down rods to resist uplift. While effective in terms of basic connectivity, its connection resilience (CR = 0) and design flexibility (DF = 1) indicate limited adaptability and minimal energy dissipation, with the damage localized between the fasteners and the CLT panels. IMC-5 offers a practical solution for shear transfer between vertically stacked modules, utilizing hardwood dowels and screws. However, the absence of any tensile or uplift mechanism (VC = 0) limits its use in seismic or wind-exposed regions. IMC-7 offers design flexibility (DF = 2) and effective horizontal connection; however, it lacks vertical connection and connection resilience features. IMC-4 integrates angle plates and pins, but its score reflects minimal design flexibility and an absence of energy dissipation concepts.

3.2. Manufacturing Evaluation

The reviewed connection systems were feasible for off-site production, although some introduced complexities that could challenge scalability for mass production. The results of the manufacturing evaluation are presented in

Table 6. Individual scores assigned for manufacturing evaluation. Manufacturing metrics abbreviations from Table 2; Connection ID from Table 4.

	Conn. ID	IMC-1	IMC-2	IMC-3	IMC-4	IMC-5	IMC-6	IMC-7
Manufacturing metrics	CC	1	0	1	2	2	1	2
	CI	2	2	2	0	1	1	2
	IP	2	1	1	1	1	1	2
	Total	5	3	4	3	4	3	6

and Figure 4.

IMC-7 achieved exceptional performance across all manufacturing parameters. This system relies on pre-engineered brackets installed using simple fasteners, such as flange head screws, with minimal machining or integration complexity [70]. IMC-7 has simple and regular geometry, straightforward component integration, and ease of installation, aligning with the emphasis on productivity and repetition in modular construction. The system’s ability to reduce fabrication time and its compatibility with acoustic isolation strategies further support its use in residential modular buildings. The acoustic isolation property was only offered by IMC-1 and IMC-7.

IMC-1 used steel plates and screws, fastened with elastic bearings between the modular units, to achieve a soundproofing effect. To resist the uplift forces, the connection utilized steel rods connected to steel plates that were placed on-site between the modules. These components were relatively easy to fabricate and integrate, enabling the simple and rapid preparation of the connection assembly during the off-site manufacturing process. However, it was considered a moderate number of parts required for the connection manufacturing, which reduced the connection’s performance in terms of CC.

Although the X-RAD system (IMC-3) is well-integrated into plant-level prefabrication workflows, its multipart nature and the need for precise alignment add moderate complexity to the production process. The connection combines a metal envelope and a hardwood insert, both of which must be previously assembled in the proprietary facilities. Later, it is installed in the corners of the CLT panels using self-tapping screws with a specific inclination angle. The CLT panel corners must be chamfered by using a computer numerical control (CNC) cutting machine, which implies a rigorous installation process for the connection.

IMC-5, the dowel-type connection system, used hardwood dowels and standard screws. The materials and methods required to install the dowel-type connections are widely available and compatible with timber workshop tooling. However, the installation of the timber dowels requires predrilled holes to facilitate the assembly process, which increases the complexity of component integration and the time required during the installation procedure. Despite the use of hardwood elements, its simplicity in geometry and traditional fastening techniques make it well-suited for high-volume prefabrication.

IMC-2, IMC-4, and IMC-6 provided moderate manufacturing complexity. The interlocking 3D-printed connections (IMC-2) required a more complex fabrication procedure for their main components. While the authors noted that the system is compatible with industrialization, its geometric complexity poses challenges for mass production using conventional methods. This connection requires a 3D printing process, which necessitates additional research and quality control testing before it can be industrialized for mass production.

IMC-6 includes both vertical and horizontal connection components. The vertical connection involves welded bolts and nut-secured interfaces, while the horizontal damper component includes custom-fabricated LYS plates. Though not highly complex individually, the combined integration of the curved steel plates for the vertical connection and the energy-dissipating elements makes this system slightly more labor-intensive than others. IMC-4, which uses a steel angle plate and fitted pins, showed a similar level of complexity. The proposed modular units require a precast concrete floor slab. The standardized steel profiles for the slab contain the pin-cone interface, which demands high precision in the prefabricated slab. Additionally, the steel plates with the pin for connection should be perfectly aligned at the roof of the modules to fit with the pin-cones of the adjacent module.

3.3. Construction Evaluation

The construction-phase performance of inter-modular connections, defined by ease of assembly, disassembly, and tolerance control, is a critical factor for evaluating the speed, precision, and cost-effectiveness promised by volumetric CLT modular systems. The results of the construction evaluation are presented in

Table 7. Individual scores assigned for construction evaluation. Construction metrics abbreviations from Table 2; Connection ID from Table 4.

	Conn. ID	IMC-1	IMC-2	IMC-3	IMC-4	IMC-5	IMC-6	IMC-7
Construction metrics	DfA	0	2	2	1	0	1	2
	DfD	0	2	2	0	0	1	2
	TC	0	2	2	1	0	2	1
	Total	0	6	6	2	0	4	5

and Figure 5.

Both IMC-2 and IMC-3 achieved the highest construction evaluation scores, largely due to their streamlined assembly process and excellent compatibility with modular construction methods. IMC-2, emphasized a sliding and stacking assembly method. This method was based on previous research that studied the interlocking mechanisms for modular construction [74]. The system’s connections are pre-installed off-site, allowing for rapid on-site installation with no need for additional fastening. IMC-2 guarantees self-alignment, simplifies tolerance control, and enables disassembly and module replacement by reversing the stacking sequence [58]. However, its success relies on the availability of specialized lifting systems and scaffolding frames, which could add logistical complexity in some construction contexts.

Similarly, the X-RAD connection (IMC-3) is pre-installed on panel corners in the factory using self-tapping screws, eliminating the need for additional on-site machining or fastening [61,62]. Once delivered, panels are positioned directly onto pre-anchored steel plates, which are developed as part of the same connection system and secured with only a few bolts. This minimizes construction labor and tooling demands. Moreover, IMC-3 enables efficient disassembly, making it a viable solution for relocatable or temporary buildings [30].

IMC-7 (ModuLink) also performed well in the construction evaluation category due to its pre-engineered bracket design, which supports fast on-site screw installation, and minimal alignment adjustments are required. Its modular adaptability and acoustic performance without the installation of additional components contribute to being a strong candidate for high-density residential projects in low seismic regions [52].

IMC-6 comprises a bolted vertical connection and a horizontal damper for energy dissipation, which requires additional steps during the assembly process to connect the components between adjacent modular units. This system involves multiple connections per module, requiring more site coordination than fully integrated solutions. Although the construction stage is slightly more complex than interlocking systems (e.g., IMC-2), its components are accessible for disassembly, and the connections provide sufficient tolerance control during construction.

IMC-4 has a relatively cumbersome installation sequence. Gijzen [67] reported that the construction process with this system involves an initial alignment of the angle plates before screwing them to the top of the side walls. Molds were used to align the pins of the plates to fit the cones of the pre-cast slab before the upper module was placed. To secure the angle plate to the concrete slab, drilled or glued anchors were suggested; after that, the next module can be placed to repeat the process. The use of drilled or glued anchors may cause difficulties during the reverse disassembly procedure and add an additional trade on-site to install the anchors, as well as another required inspection, thereby adding time and cost to the construction. This multi-step process also introduces greater potential for alignment issues and on-site delays.

IMC-1 and IMC-5 have substantial limitations in their construction-phase performance. IMC-1 involves multiple screw types and faces accessibility challenges, especially for vertical connections when modules are stacked next to each other, with only one module vertically connected, while the second module remains disconnected. These issues complicate installation and tolerance control, making the connection less suitable for rapid modular deployment. Similarly, IMC-5 lacks features that would facilitate module positioning, and it does not allow for disassembly. This connection does not incorporate alignment aids or tolerance adjustment methods, which are important for efficient on-site stacking.

3.4. Experimental and Numerical Evaluation

Experimental validation and numerical modeling are crucial for verifying the mechanical behavior, particularly at large deformations, the failure mechanisms, and reliability of inter-modular connection systems in CLT modular construction. These investigations provide an empirical and computational basis for understanding load transfer mechanisms, deformation modes, ductility, and failure sequences. The experimental tests were also used to calculate the strength and stiffness of the inter-modular connections, which defined numerical or analytical models for the design of modular structures. The results of the experimental and numerical evaluation are presented in

Table 8. Individual scores assigned for experimental and numerical evaluation. Experimental and numerical metrics abbreviations from Table 2; Connection ID from Table 4.

	Conn. ID	IMC-1	IMC-2	IMC-3	IMC-4	IMC-5	IMC-6	IMC-7
Experimental and numerical metrics	CET	0	1	1	0	1	0	0
	FSET	0	1	1	0	0	1	1
	NME	1	2	2	1	2	0	0
	AMD	0	1	1	0	1	0	1
	Total	1	5	5	1	4	1	2

and Figure 6. The assigned scores were relatively low across most systems, suggesting that few of the reviewed connections were experimentally or numerically explored, or that the data and analysis from such investigations was not published.

IMC-2 and IMC-3 lead this category with the highest scores, supported by detailed experimental testing and numerical simulations. IMC-2 was tested under both monotonic and cyclic loading of shear and tensile connections. The tests revealed key damage limit states, including buckling of the steel plates and bending of L-shaped male connections. Furthermore, sophisticated finite element models were developed using orthotropic elastoplastic timber and hardening steel models, capturing realistic connection behavior under various loading regimes. An important modeling methodology to capture the screw withdrawal effect [75] was incorporated to simulate the weakening effect of fasteners and verify the existence of localized timber damage.

IMC-3 was extensively explored by Angeli et al. [60], Polastri et al. [61,62], and Bhandari et al. [64,65,66]. This connection has been widely tested through both component-level and full-scale experimental setups. Testing covered various loading angles and panel configurations, revealing detailed performance parameters such as screw withdrawal resistance, ductility factors (ranging from 2.0 to 6.8 depending on direction), and failure modes involving screw fracture and wood cracking. Numerical analyses were conducted using both SAP2000 link elements and more advanced OpenSees two-node link element models to simulate directional stiffness and load-bearing behavior [62,65]. These models were benchmarked against experimental tests of monotonic and cyclic loading performed with different CLT species.

IMC-5 was investigated experimentally through 96 tests on single-shear CLT joints using hardwood dowels and screws. The study systematically examined fastener spacing, grain orientation, and fastener diameter using both monotonic and cyclic loading protocols. While this testing provides substantial insight into material behavior, the evaluation focused on generic CLT-to-CLT joints rather than the behavior of the connections within the volumetric module. Therefore, the application to a volumetric module seems tenuous. Although experimental testing of IMC-7 was performed [73], no numerical validation was reported to support the design process of this connection. The only information provided for this connection system corresponds to the published data sheet, which includes stiffness and strength values for design purposes [70].

The remaining systems, IMC-1, IMC-4, and IMC-6, had minimal or no published experimental or numerical information. For instance, IMC-1, despite being described by Koskimies [55] as structurally viable, lacks experimental test data or simulation results. Similarly, IMC-4 (angle plate design by Gijzen [67]) and IMC-6 (damper-based connection by Zhang et al. [69]) were presented conceptually or with limited experimental evaluation. Zhang et al. [69] performed reverse cyclic tests on wood assemblies that included the connection; however, without quantified performance metrics or accessible modeling details, the validation remains incomplete.

As a complement to the experimental and numerical evaluation,

Table 9. Comparison of experimental results from inter-modular connections.

ID	Experimental Test	Loading Protocol	Damage	Failure Modes
IMC-2	Tensile connection: Three monotonic tests Two cyclic tests Shear connection: One monotonic test (two specimens) One cyclic test (two specimens)	Monotonic loading (0.05 mm/s) Cyclic loading (0.02 mm/s) EN 12512 [76] with estimated yield point of 4 mm in tension and 2 mm in shear	Tension Shear	Plastic deformation within the male connection Tensile connection [58]: Bending of the L-shaped elements in the male connection Shear connection [58]: There was a sudden drop in the force after reaching 4 mm, followed by another after reaching 6.5 mm, indicating buckling in the shear connection
IMC-3	Preliminary tests for screws withdrawal capacity The connection was tested in five different loading configurations as described in [60,61]	Monotonic loading (0.1 mm/s) EN 26891 [77] Cyclic loading EN 12512 [76]	Tension Shear	Tension configuration [61]: Block-shear failure, local failure in the hardwood insert Shear, Tension-shear, and compression-shear [61]: Tensile failure of screws, cracks and splitting of wood insert
IMC-3	Shear-tension and shear compression monotonic tests Wall systems assembled with X-RAD	Monotonic loading (0.1 mm/s) EN 26891 [77] Cyclic loading EN 12512 [76]	Tension Shear Tension-shear Compression-shear	Tension [62]: Block tearing of the metal envelope Shear [62]: Tensile rupture of screws Tension-shear and Compression-shear [62]: Tensile rupture of screws
IMC-3	Tests using Ponderosa Pine CLT Monotonic test specimens: Two specimens in tension One in compression One in tension-shear One in compression-shear Cyclic test specimens: Three in cyclic tension-compression Four in cyclic shear-tension and shear-compression	Monotonic loading EN 26891 [77] Cyclic loading EN 12512 [76]	Tension/Compression-shear Tension	Tension-shear [66]: Tensile fracture of the screws Compression-shear [66]: Shear fracture of the screws at the connection-CLT interface Tension-shear and compression-shear [66]: Yielding of the screws in tension loading followed by shear fracture of the screws in compression loading Vertical cracks in the hardwood insert
IMC-5	Monotonic and cyclic single-shear plane CLT-to-CLT joint tests using two species of hardwood dowel (RO, B) [68] and large-diameter screws installed at 90- (S90) and 45-degree angles (S45) (tension and compression for 45°)	Monotonic loading ASTM D5764 [78] Cyclic loading ASTM E2126 [79] CUREE basic protocol	Hardwood dowels Screws	RO and B [68]: Longitudinal shear crack in the hardwood dowels S90 [68]: Two hinges in the screw corresponding to Mode IV, according to NDS [80] S45 [68]: Withdrawal and buckling when loaded in positive and negative phases, respectively
IMC-6	Five cyclic tests for the vertical connection Three cyclic tests for the horizontal connection	Modified CUREE basic loading protocol (0.5 mm/min)	Horizontal connection	Vertical connection [69]: Permanent deformation of top plate Horizontal connection [69]: Stress concentration occurred at the corners of the low-yield steel dampers
IMC-7	Each test was executed five times: The first test to determine Fmax (Fmax deformation = 15 mm), then four monotonic tests. Shear tests with four connections each, compression tests with four connections each, and tensile tests with two connections each. [73]	Monotonic loading EN 26891 [77] (3 mm/min)	Shear Compression Tension	Shear [73]: Pronounced shearing of the spacing elastomer, rotation of L-plates and slip of the screws. Compression [73]: Complete compression of the elastomeric bearing. Yielding of L-plates. Contact between bolts and timber, indentation in the timber. Tension [73]: Compression of the elastomeric bearing, yielding of the L-plates, and pronounced bending of the bolts.

summarizes the experimental tests performed by previous researchers on the inter-modular connections. This table includes information related to loading protocols used for their experimental testing, an illustration of the reported damage, and the failure modes detected during the experiments. From the connections described in Table 4, some cases remained as theoretical projects and lacked experimental or numerical tests to support the design.

From the reported observed damage to the inter-modular connections from the experimental investigations listed in Table 9, some of the experimental investigations resulted in damage localized in the CLT panels. Tests on IMC-3 and IMC-5 resulted in crushing of the wood in contact with the screws and hardwood dowels, respectively. Testing of IMC-3 resulted in yield and fracture of the screws due to a combination of tensile and shear stresses, while testing of IMC-5 resulted in withdrawal and buckling of the screws during the cycling loading tests.

Conversely, no significant damage to timber panels was observed during testing of IMC-2, IMC-6, and IMC-7. The design of IMC-2 ensured damage within the connection and not the timber. This was achieved through bending of the L-shaped elements in tension and buckling of the curved plates in shear. IMC-6 showed yielding of the dampers provided for the horizontal connection. Similarly, testing of IMC-7 exhibited deformation of the elastomeric bearing elements, localized between the L-plates of the connection, in addition to yielding of the L-plates under both tensile and compressive loads. The significance of the location of damage can be in the recovery of the building. For connections where the damage was localized within the timber itself, the entire CLT panel will need to be replaced within the volumetric unit. In addition, energy dissipation can be assumed to be in the timber, which can result in brittle failure modes. However, for connections where the damage is located in the connection itself, in the recovery of the building, only the connection would have to be replaced. In addition, the energy dissipation is from the steel components within the connection, which is ductile and can withstand large deformations prior to fracture.

3.5. Overall Ranking

The overall ranking of the evaluated inter-modular connections provides a synthesized performance comparison across all categories: structural, manufacturing, construction, and experimental and numerical investigation.

Table 10. Summary of assigned scores per category to each inter-modular connection.

Conn. ID	IMC-1	IMC-2	IMC-3	IMC-4	IMC-5	IMC-6	IMC-7
Structural	3	6	5	2	3	4	3
Manufacturing	5	3	4	3	4	3	6
Construction	0	6	6	2	0	4	5
Exper. & Num.	1	5	5	1	4	1	2
Total	9	20	20	8	11	12	16

provides the individual scores per category for each connection, and Figure 7 shows the cumulative scores for the seven connections described in Table 4. The results span from 8 (IMC-4) to a maximum of 20 (IMC-2 and IMC-3), reflecting variability in the design and implementation of the inter-modular connections.

IMC-2 and IMC-3 had the highest overall scores. IMC-2 performed consistently well across all categories. Its high structural score reflects a well-integrated design with interlocking tension and shear mechanisms, designed to localize damage in the connection rather than the timber panels and provide easy disassembly. Despite its geometrical complexity, the system’s industrial compatibility and factory-preinstalled nature resulted in strong performance in the manufacturing and construction evaluations. The connection’s extensive experimental validation and nonlinear numerical modeling further supported its robustness and reliability.

Similarly, IMC-3 demonstrated good performance across all evaluation categories. Its modular, corner-based configuration and standardized design contributed to high construction scores, particularly due to the ease of lifting, positioning, and anchoring on-site. Experimental programs, ranging from component-level testing to full-scale testing, coupled with analytical and numerical models, provided comprehensive validation of the system’s performance. Although installation requires specialized technology, such as CNC machining for the CLT panels, these are compatible with high-precision prefabrication environments.

IMC-7 had high ranking in manufacturing performance and high construction efficiency, due to its simple design and reduced labor-intensive on-site operations during installation. However, due to the lack of vertical connectivity and the absence of numerical simulations published in the literature, the connection exhibited lower structural and experimental performance. IMC-5 and IMC-6 achieved intermediate scores. IMC-6 integrates vertical and horizontal components with energy dissipation mechanisms in the horizontal direction, showing design objectives for seismic applications. However, its experimental evaluation remains limited, and construction complexity slightly reduces its cumulative score. IMC-5 benefits from straightforward manufacturing using standard fasteners and dowels and includes robust cyclic test data for fastener performance. Still, it lacks the vertical load transfer capacity and features necessary to support modular disassembly or alignment, thereby limiting its applicability as a comprehensive solution for modular construction.

IMC-1 and IMC-4 had the lowest overall scores. Both inter-modular connections were conceptually designed to meet the demands of CLT modular buildings; however, they lack of experimental testing and a suitable construction methodology. IMC-1 presents installation challenges due to its limited accessibility and the intensive use of screws in assembly. IMC-4 showed relatively low performance in all categories. Both connection systems’ reliance on precise alignment and multi-step anchoring procedures, coupled with the absence of energy dissipation mechanisms and supporting test data, constrains their adaptability for current modular construction needs, particularly in regions of high seismicity or wind.

4. Discussion

Volumetric modular buildings are still an emerging construction methodology that lacks a defined design philosophy to ensure a predictable response from all modular structures under lateral loading. The design process usually follows traditional lateral design methods [81,82,83] and includes performance objectives [14] to satisfy the structural demands. The identified volumetric CLT modular buildings (Table 3) and the published data on inter-modular connections revealed limited existing knowledge for design and evaluation of the connections in regions with high seismic and wind demands.

The structural evaluation revealed that all identified inter-modular connections provided horizontal connectivity, fundamental to guarantee the horizontal load transfer between the modules to ensure a diaphragm behavior [14]. However, not all the connections provided vertical connectivity to meet the uplift demands present in intermediate and high seismic regions, creating discontinuities in the load path. On the contrary, other approaches, such as interlocking mechanisms [58,74,84], combined the horizontal and vertical connectivity and improved the stability of multi-story modular buildings by generating additional redundancy and alternative load paths between the modular units. The lack of a continuous load path in modular buildings can introduce stress concentrations, localized damage, or increase the potential for failure [85,86].

Additional objectives for improved structural performance included connections with energy dissipation capacity [58,69] and connections that provided design flexibility to accommodate the modules into different building layout designs. The building configuration and connection design configuration play an important role in defining the failure mechanisms expected, providing structural redundancy and resilient solutions that concentrate damage in specific components or delay global damage of the buildings. Li et al. [87] identified that 64% of modular buildings follow a rectangular-based shape layout, while the remaining 36% included non-rectangular layouts. Therefore, the ability of these connections to be adaptable to various layouts provides a larger market for their use within buildings.

For the design of the connections, traditional strength design of the components is usually performed to satisfy the demand requirements. There are no available standards for modular construction. Based on the current standards for timber structures [80,88,89], there is a gap in standardized requirements for the design of intermodular connections; therefore, the design is based on the strength of the desired connection configuration to satisfy the lateral demands in the modular buildings.

The manufacturing evaluation suggests that most of the connections are still within a theoretical or development stage. The evaluation revealed a lack of reported quality control methodologies, a lack of testing methods needed for the industrialization process, and minimal or no information published about the fabrication procedure of the connections. However, despite the lack of data, IMC-3 and IMC-7 provided installation manuals to facilitate their installation procedure, following the designer specifications [63,71].

Additional information from off-site procedures is required to improve the evaluation performance for the assembly and installation process of the connections. Similarly, to improve the manufacturing processes, Tenório et al. [11] highlighted the necessity for a continuous development of BIM tools in the AEC industry for timber products and their interconnection with machinery software to facilitate in situ production (e.g., including the design of cuts and holes in timber panels).

The construction performance evaluation revealed a lack of building examples where the connection systems had been applied. IMC-7 was the only connection with reported case studies [52]; however, they did not include details from the construction process. IMC-3 provided full-scale prototypes that are recommended to document the assembly and disassembly process with an inter-modular connection system to obtain the required time, equipment, and number of workers needed to complete the task [90]. IMC-2 reported a well-detailed assembly and disassembly process [58], supported by conceptual studies to support the proposed construction methodology [74,84]. These limited cases in the construction methodologies for modular buildings revealed a high degree of uncertainty in selecting a connection system and a lack of study cases to support new designs of modular buildings. The construction performance can be translated into cost and time savings during the modular building’s construction phase. When a connection system provides a clear construction method, on-site delays are minimized.

The performance score of a given connection across each category could be used to identify flaws or weak characteristics of a connection system during the design process. Based on Shan and Pan [91] methodology for the design of modular buildings, the proposed manufacturing and construction evaluation could enhance the modular system schematization and the preliminary design of a modular building. Similarly, structural, experimental, and numerical evaluations could provide additional support before analyzing the global and local performance of the building’s components.

The testing methodologies required to evaluate inter-modular connections subjected to high lateral demands are highly dependent on the connection configuration and the plasticity mechanisms of the connection system. Monotonic and cyclic loading protocols are recommended to evaluate the response [3], focusing on parameters such as stiffness, strength, ductility, and energy dissipation. Based on the identified connections, different testing methods, performance requirements, and loading protocols were used. A consistent testing methodology should be developed for these connections to evaluate them under similar conditions. In addition, full-scale building testing [92] should be performed to benchmark component testing and determine whether it is representative of the building’s overall performance. In parallel, numerical models can be validated against experimental data to generalize the connection response and evaluate the global performance of the entire structure.

5. Future Research

To promote the construction of volumetric CLT modular buildings in regions with high seismic and wind hazards, improvements in the connection systems are required to meet the additional structural demands. Further evaluation with different layout configurations is necessary to predict the response of volumetric CLT modular buildings. To fulfill these requirements, a design methodology is needed to standardize the analysis of modular buildings and identify the structural demands that govern their response.

Extensive experimental campaigns were performed for some of the identified inter-modular connections (Table 9), followed by numerical evaluations with published data from a few of them. Unfortunately, none of the previous studies performed an analysis of the global response of a modular building based on the behavior of the proposed connections. These studies focused their attention solely on the local performance of the connection, developing analytical and numerical models without further development. Evaluation of the global response of the building can be used to assess and verify the performance objectives defined during the design, and include the dynamic response of the buildings up to collapse stages, similar to previous studies on steel modular buildings [93,94]. In addition, numerical simulations of the full building would provide data to develop a design guide for these buildings that could clarify performance objectives and the desired location of energy dissipation. Previous researchers have developed design methodologies and workflows for concrete and steel modular buildings [91,95,96]. To follow a similar approach, a set of performance objectives should be established based on the global performance of CLT modular buildings and complemented with the required local performance of the selected inter-modular connection. These types of design guides can also provide industry-supported standards that can assist Authorities Having Jurisdiction (AHJs) in understanding the quality of designs and thereby enabling more streamlined building reviews.

The scores (Table 10) from the construction performance category presented the largest variability in the evaluation of the connections, followed by the experimental and numerical evaluation category. This variability in performance highlights the absence of detailed construction methodologies developed for modular workflows and cases that demand complex operations, time-intensive processes for installation, or intricate future disassembly operations of the modular buildings. Similarly, it reveals a lack of experimental tests or reported data to provide sufficient support for the application of these connection systems.

The manufacturing of the connections requires off-site evaluations to measure the performance of the fabrication procedures. Zhang et al. [97] proposed a process-oriented framework to improve production methodologies applicable to the modular construction industry. However, it requires an initial collection of data to apply the method. With the execution of experimental tests and fabrication of modular prototypes, statistical data should be collected to identify the components or activities that add higher complexity and increase the time required for connection installation. In addition, further data on off-site construction would provide information on tolerance limits required for the buildings and their influence based on the connection system configuration (e.g., connection spacing, accessibility). The data from these off-site construction methods can provide insight into where new technologies are most beneficial to increase speed of construction. For example, laser scanning methods [98] can incorporate solutions for construction tolerances, reducing initial eccentricities of modular buildings due to inadequate installation alignment.