Assessment of UHPC with Various Particle Distributions (q) and Low Cement Consumption

Abstract

Ultra-high-performance concrete (UHPC) has been increasingly adopted in applications requiring superior mechanical performance and high durability under aggressive environments. However, its large-scale use is still limited by the high binder content and the lack of a standardized mix design methodology. Among the existing approaches, particle packing-based mix design methods have shown the most promising results, optimizing the composite structure and enabling efficient material proportioning. This study aimed to evaluate the influence of the particle distribution coefficient (q = 0.20 and 0.25) and the cement consumption ratio (15%, 20%, and 25%) on achieving the lowest packing deviation index (PDI) values using a rational UHPC mix design method. The results indicated that increasing q allowed a reduction of up to 15% in cement content, corresponding to 106 kg/m³ less binder. In contrast, changes in cement consumption, which led to different PDI values for the same q, had a significant effect on compressive strength. Mixtures with 20% cement and consumption of 598 kg/m³ exhibited the lowest PDI values (180 and 190) and the highest 91-day compressive strengths (147.0 and 151.1 MPa). Fiber reinforcement improved toughness and post-elastic energy absorption capacity. Overall, UHPC with reduced cement content and high mechanical performance can be achieved using a rational mix design method when an appropriate q value is selected.

1. Introduction

Ultra-high-performance concrete (UHPC) is one of the most innovative cementitious composites developed in the past three decades, offering superior strength, durability, and workability compared to conventional concrete [1]. According to Rahhem, Mahdy and Mashlay [2], UHPC has gained prominence due to the increasing performance demands in modern construction, exhibiting compressive and flexural strengths exceeding 150 MPa and 10 MPa, respectively, as well as high ductility, toughness, and enhanced resistance to aggressive agents.

Given these characteristics, UHPC is well-suited for special structures requiring high load-bearing capacity or protection against the ingress of harmful substances, such as chloride ions and carbon dioxide, as well as for extending service life in harsh environments involving freeze–thaw cycles [3]. The outstanding performance of UHPC originates from its extremely low porosity [4], which results in a denser matrix structure compared to conventional concrete. This low porosity corresponds to a high particle packing density [5,6]. Consequently, UHPC has been increasingly used in various types of projects, ranging from marine and underground works to supertall buildings and other special applications [7,8].

Malik et al. [9] conducted an extensive survey of UHPC-related publications and reported an average cement consumption of 789.13 kg/m³, with a standard deviation of 199.52 kg/m³, highlighting the significant variability among different studies. According to the authors, cement consumption is strongly influenced by particle packing, and substantial reductions in cement content can be achieved while maintaining high mechanical performance through optimized granulometric design. Such optimization enables more efficient structural designs with reduced concrete volume, thereby lowering the carbon footprint when compared to structures designed using conventional 60 MPa concrete.

In search of more sustainable UHPC formulations, several studies have investigated the partial replacement of cement with alternative materials such as ground granulated blast furnace slag, fly ash, and other supplementary cementitious materials, as well as the development of rational mix design methods [10,11,12,13].

Regarding UHPC mix design, there is still no consensus on an effective mixing procedure, as numerous parameters are used across different methodologies, hindering large-scale implementation [1]. As stated by Ahmed et al. [14], a proper UHPC mixing procedure is essential to obtain a uniform mixture. An efficient mix design method not only improves uniformity but can also reduce mixing time and overall production costs [15]. These considerations highlight the need for systematic approaches that can optimize both material properties and production efficiency.

In this context, recent studies have emphasized the relevance of systematic UHPC mix design, particularly with respect to particle packing and hydration, demonstrating that such approaches are crucial for further optimizing UHPC production. Physical-chemical packing methods improve particle size distribution and microstructure, enhancing density and mechanical properties [16]. Similarly, optimization of cementitious materials and supplementary additions has been shown to benefit workability and flexural performance, reinforcing the importance of rational mix design strategies [17]. Moreover, mathematical methods, such as response surface methodology and multi-objective evolutionary algorithms, support the definition of mixture proportions that maximize both performance and sustainability [18], while adaptations of classical packing models, such as Andreassen, effectively guide the development of dense, high-performance mixtures [19].

UHPC mix design studies generally rely on particle packing density and particle size distribution (PSD) models [20]. Among these, two approaches are widely used: the semi-empirical continuous PSD model proposed by Andreasen and Andersen [21] and later refined by Funk and Dinger [22], and the Compressible Packing Model (CPM) developed by Larrard [23]. These models provide a theoretical framework for designing UHPC mixtures with optimized packing and reduced porosity.

Building on these concepts, Christ [24] developed a rational mix design method based on the particle packing curve, introducing the Packing Deviation Index (PDI) as a quantitative measure of how closely a mixture follows the ideal packing curve. The PDI is calculated as the sum of the areas between the theoretical particle packing curve (from Funk and Dinger [22], modified from Andreasen and Andersen [21]; see Equation (1)) and the actual mix curve derived from the particle size distributions and proportions of each material, as expressed in Equation (2) [24]. This approach provides a practical metric for evaluating and optimizing UHPC mixtures.

$${\displaystyle \frac{PSD}{100\%}}={\displaystyle \frac{D^q-D_{min}^q}{D_{max}^q-D_{min}^q}}$$

(1)

$${\displaystyle \frac{CPP}{100}}=\sum \left(Q\times D_{retained}\right)$$

(2)

where PSD represents the particle size distribution, D_min and D_max are the minimum and maximum particle diameters considered, respectively. The factor q is the particle distribution coefficient, which ranges from 0 to 1; the lower the q value, the greater the proportion of fine particles in the mixture. Furthermore, CPP is the cumulative passing percentage, Q is the percentage of each material in the mixture, and D_retained is the percentage retained at a given particle size.

This method is based on achieving the best possible particle packing using the materials available for UHPC production. This condition is represented by PDI values approaching zero, indicating a mix curve closer to the ideal packing [25].

The rational mix design method proposed by Christ [24] is promising because it allows the preliminary determination of UHPC mixtures prior to experimental testing. Based on the PDI, it is possible to predict the packing efficiency of a mixture relative to the ideal packing curve. In this context, the present study aims to evaluate the influence of the particle distribution coefficient and different cement contents using a rational mix design method, thereby contributing to the development of a more broadly applicable rational mix design approach. The analysis focuses on the effects of the particle distribution coefficient (q) and the cement consumption ratio on mixture workability, as well as their impact on the compressive and flexural strengths of UHPC.

2. Experimental Program

2.1. Materials

The mix design method employed prescribes that the selection of materials should aim to identify those locally available and to analyze aspects such as the mineral origin in order to assess hardness, particle size, and chemical composition. According to Christ [24], this step is essential because it requires selecting materials with distinct particle size ranges to avoid overlapping granulometries.

The materials selected for this research were high-early-strength Portland cement, silica fume, quartz powder, fly ash, and washed river sand. The chemical and physical characterization tests were performed, and the results are presented in

Table 1. Chemical composition of the materials.

Material	Chemical Composition (%)
	SiO2	Al2O3	Fe2O3	CaO	MgO	SO3	K2O	Na2O	PF
High-Early-strength Portland Cement	18.54	4.51	2.75	62.09	3.62	2.85	0.43	0.12	3.40
Silica fume	95.09	-	-	-	-	-	-	0.45	2.03
Fly ash	60.23	20.71	10.56	2.32	1.57	0.86	1.97	-	0.37
River Sand	>99.00	-	-	-	-	-	-	-	-

and

Table 2. Physical characteristics of the materials.

Material	Specific Mass (g/cm3)	Specific Surface Area (m2/g)	Mean Particle Size (µm)
High-Early-strength Portland Cement	3.09	4.40	13.36
Silica fume	2.08	3.079	1.00
Fly ash	2.10	3.519	30.52
River Sand	2.70	-	210.7

, respectively.

The cement properties that influence UHPC performance are primarily the C₃A content and the specific surface area. Meeting these requirements, a high-early-strength Portland cement with a characteristic compressive strength of 34 MPa at 7 days was used. The silica fume employed in this study was selected due to its high fineness and non-densified nature, which favors better dispersion of the fine particles within the UHPC mixture. The fly ash, another pozzolanic material, was obtained from coal combustion in a thermoelectric power plant. The sand used in this research is a quartz-based river sand.

The superplasticizer admixture (HRWR) used was a polycarboxylate ether (PCE)-based superplasticizer with high performance. The metallic fiber incorporated in the mixes was copper-coated, with a diameter of 0.21 mm and a length of 13 mm, exhibiting a tensile strength of 2750 N/mm² and an elastic modulus of 200 GPa.

It is essential to determine the particle size distribution of the materials used, as this information is required for the application of the mix design method. For this purpose, laser granulometry tests were performed to determine the retained percentages for each particle diameter of the materials. For sand, which has coarser particles, the sieve analysis method was applied. Figure 1 presents the particle size distribution curves of all materials used in this study.

It can be observed that, among the materials used in this study, the cement and fly ash exhibit very similar particle size ranges, with cement being slightly finer. However, the fly ash shows a more discontinuous particle size distribution, which contributes to improved particle packing [25].

Table 3. D90, D50 and D10 of the UHPC mixtures.

Mix Design	D90 (µm)	D50 (µm)	D10 (µm)
173-15%-0.20	266.38	72.57	0.66
180-20%-0.20	267.12	71.39	0.83
203-25%-0.20	268.79	71.99	2.09
204-15%-0.25	273.15	105.53	2.10
190-20%-0.25	273.06	101.05	3.57
214-25%-0.25	275.04	109.90	5.42

presents the D90, D50, and D10 diameters of the mixtures to highlight the nuances in the variation in the Packing Deviation Index (PDI) and the particle distribution coefficient (q). A reduction in the distribution coefficient q to 0.20 results in a smaller D50 and a higher proportion of fine particles (D10), whereas q = 0.25 leads to coarser particles for both diameters, reflecting a broader particle size distribution curve.

2.2. Concrete Mix Design

With the materials defined, the second stage consisted of determining the proportions of each component in order to obtain the lowest packing deviation index (PDI) for the mixtures. Since the materials themselves were not changed, adjustments to the PDI values required varying the material proportions. Therefore, the cement content was modified to 15%, 20%, and 25%.

For each cement percentage, the particle distribution coefficient (q) of the modified A&A model (Equation (1)), proposed by Funk and Dinger [22], was also varied between 0.20 and 0.25 in order to evaluate the influence of this parameter on both the fresh and hardened states.

Table 4. Summary of the mix proportions and quantities per m³.

Mixtures	Mass Composition Per Cubic Meter (kg/m3)
	Ce-Ment	Silica Fume	Fly Ash	River Sand	SP *	Water	Fiber Content
173-15%-0.20	448	547	390	1141	27.7	277	50.5
180-20%-0.20	598	467	348	1157	28.2	283	51.4
203-25%-0.20	747	383	296	1189	28.5	285	52.3
204-15%-0.25	448	383	448	1274	25.6	256	51.1
190-20%-0.25	598	350	370	1275	26.4	264	51.9
214-25%-0.25	747	287	284	1322	26.4	264	52.8

* SP: superplasticizer admixture.

summarizes the six proposed UHPC mixtures, where the first number corresponds to the mixture PDI, the second to the cement content, and the third to the particle distribution coefficient. The water/binder ratio (w/b ratio) was kept constant at 0.20 for all mixtures. Figure 2 shows the particle size distribution of the mixtures compared to the ideal packing model distribution.

2.3. Maximum Fiber Content

The maximum fiber content to be incorporated into the mixture was determined using the modified equation initially proposed by Martinie, Rossi and Roussel [26], as presented in the method by Christ [24]. The volumetric fraction of sand in the mix design and the dense packing fraction of the mixture was adjusted considering the bulk density (1520 kg/m³) and the density of the sand (2700 kg/m³), respectively, in order to more accurately represent the actual behavior of the material within the mixture (Equation (3)). Considering the aspect ratio of the steel fiber (r = 62), the resulting value was 2.82%, and a value of 2.50% was adopted.

$$T_{fibers}={\displaystyle \frac{400}{r}}\times \left(1-{\displaystyle \frac{M_{bulkdensity}}{M_{density}}}\right)$$

(3)

where T_fibers represents the fiber content incorporated into the mix, r corresponds to the fiber shape ratio (the ratio between the fiber length and thickness), M_bulk density denotes the bulk density of the sand, and M_density refers to its specific density.

2.4. Sample Preparation

The mixing process followed the same sequence for all mixtures, maintaining identical material addition order, mixing time, and paddle rotation speed. A planetary mixer with a 300 L capacity and variable-speed control (0–100 rpm) was used for all batches. The mixing sequence began with the homogenization of dry materials for 1 min, followed by the addition of the total amount of water and then the entire dosage of the superplasticizer. The mixtures were then mixed for approximately 5 min at 10 rpm and subsequently for another 5 min at 50 rpm. The steel fibers were incorporated over a period of 2 min at 10 rpm, and the final homogenization was carried out for 3 min at 50 rpm.

2.5. Testing Procedures

All mixtures were evaluated in the fresh state by measuring the consistency spread diameter obtained by the flow table test, according to ASTM C1611/C1611M-18 [27]. The flow diameter was determined as the average of two perpendicular measurements of the circular spread formed by the UHPC after flow ceased. The flow time was recorded from the moment the slump cone was lifted until the mixture reached a diameter of 600 mm on the base plate. Additionally, the visual stability index was assessed to identify the occurrence of segregation or bleeding by observing the formation of central mounds and peripheral halos of fines on the spread surface.

In the hardened state, the mixtures were evaluated for compressive strength, flexural strength and toughness factor. For each mixture and testing age, five specimens were cast and tested for compressive strength in accordance with ASTM C39/C39M-20 [28]. Cylindrical specimens (100 × 200 mm) were cured under moist conditions (25 °C ± 3 °C and 95% RH) until testing at 7, 28, and 91 days, using a servo-hydraulic testing machine with a 2000 kN capacity. For the evaluation of flexural strength and toughness factor, two prismatic specimens per mixture and testing age were tested in accordance with ASTM C1609/C1609M-19 [29]. Prismatic specimens measuring 150 × 150 × 500 mm were tested at 28 and 91 days.

3. Results

3.1. Fresh Properties of UHPC Mixtures Analysis

According to the results presented in

Table 5. Flow test results for UHPC mixture.

Mixtures	Water/Binder Ratio	Spread (mm)		Spread Classification (Recommended Practice NBR 17246, 2025)
		B1	B2
173-15%-0.20	0.20	780	800	SF3
180-20%-0.20	0.20	660	640	SF2
203-25%-0.20	0.20	660	650	SF2
204-15%-0.25	0.20	625	620	SF1
190-20%-0.25	0.20	810	800	SF3
214-25%-0.25	0.20	650	635	SF1

, the mixtures with a particle distribution coefficient q of 0.25 exhibited an average spread of 690 mm, whereas those with q = 0.20 showed a slightly higher average value of 698.3 mm. This behavior indicates that lower q values tend to provide greater flowability, possibly due to a more efficient particle packing that reduces the internal friction within the mixture.

Furthermore, within each group of q values, it was observed that mixtures with lower PDI values resulted in greater flow diameters, indicating that a lower cement content enhances the workability of the material under the same w/b ratio for the mixtures evaluated. This behavior may also be related to particle morphology. According to Neville [30], cement grains, which are predominantly lamellar and irregular in shape, tend to reduce particle mobility. In contrast, silica fume and fly ash, with finer and more spherical particles, promote better dispersion and flowability, thus improving the spread performance.

3.2. Compressive Strength Analysis

According to Figure 3, it can be observed that increasing the cement content from 15% to 25% resulted in higher compressive strengths at early ages (7 days), a behavior similar to that reported by Ahmed et al. [14] and Lv et al. [31], who also observed increased strength in mixtures with higher cement consumption. According to Wu, Shi and He [32], the presence of fly ash tends to reduce compressive strength up to 7 days under standard curing conditions. Consequently, mixtures with higher fly ash contents (PDI 204, 190, 173, and 180) exhibited the lowest compressive strength values.

In matrices containing supplementary cementitious materials (SCMs), early-age strength development is predominantly governed by the cement content. The SCMs react secondarily with the hydration products of cement, contributing significantly to strength gain at later ages [33].

With curing age progression (28 and 91 days), an expected increase in compressive strength was observed for all mixes. However, an inverse trend was detected among the mixtures. Mixes with lower cement contents (15% and 20%) exhibited strength gains of up to 30% at 28 days, contrary to the behavior observed at early ages. The cement content of 20% resulted in the highest compressive strength at both 28 and 91 days, indicating it as the optimal cement dosage within the parameters of this study. The mixture PDI 190-20%-0.25 achieved the best mechanical performance among the evaluated compositions, reaching a compressive strength of approximately 151 MPa after 91 days of curing.

Two main factors may explain the lower strength observed in mixtures with 15% cement content. The first concerns the presence of pozzolanic materials with particle sizes exceeding 45 µm, which, according to Mehta and Monteiro [33], do not participate in secondary pozzolanic reactions, thus limiting the formation of hydration products. In mixtures PDI 204 and PDI 173, the fly ash content was 20.3% and 17.6%, respectively, with approximately 25% of the particles above this cutoff. The second factor is related to the low cement content combined with a high amount of silica fume, which may reduce strength due to the limited availability of calcium hydroxide required for secondary reactions, consequently affecting the strength development [34].

When assessing the influence of the particle distribution coefficient (q) among mixtures with the same cement content, no significant differences in compressive strength were observed at any curing age. This behavior indicates that small variations in q have no substantial impact on this property. A similar trend was reported by Liu et al. [35], who found no significant effect on the compressive strength of UHPC mixtures for q values ranging from 0.21 to 0.27, keeping the same maximum particle size (D_max).

As curing age advanced, however, mixtures with lower PDI values exhibited higher compressive strengths. The mixtures PDI 190 and PDI 180 achieved the highest results, with strengths of 138.3 and 151.1 MPa and 134.5 and 147.0 MPa at 28 and 91 days, respectively. Lv et al. [31] also observed that denser UHPC mixtures containing fly ash and silica fume exhibited higher compressive strengths, consistent with the findings of the present study.

In contrast, the mixtures PDI 204 and PDI 173 presented the lowest strengths at both curing ages. Comparing the particle packing behavior, the PDI 173 mixture exhibited slightly higher strength than PDI 204 at 91 days, suggesting that, for identical cement contents but different PDI values, a lower PDI and improved packing efficiency result in denser and stronger matrices.

Another aspect to be considered is that, although the water-to-binder ratio was maintained constant for all mixtures, variations in the cement content relative to the total binder resulted in different water-to-cement ratios. This change may affect cement hydration kinetics and early-age strength development. Nevertheless, in UHPC systems composed of multiple fine constituents, the water-to-binder ratio is commonly adopted as the governing parameter, since supplementary cementitious materials significantly contribute to particle packing, pore refinement, and matrix densification.

In this context, the mechanical performance observed cannot be attributed solely to cement hydration but rather to the combined effects of optimized particle packing, improved distribution of the available mixing water, and the synergistic action of cement and pozzolanic materials. While the influence of water-to-cement ratio cannot be completely decoupled from these mechanisms, the results indicate that particle packing optimization enables a more efficient use of cement, allowing reductions in cement content while maintaining comparable mechanical performance. Future studies focusing on microstructural characterization and hydration kinetics could further isolate and quantify these individual effects.

To verify whether statistically significant differences occurred among the mixtures at 7, 28, and 91 days, an analysis of variance (ANOVA) was performed (

Table 6. ANOVA of compressive strength at 7, 28 and 91 days.

		G.L	Sum of Squares	Mean Square	F-Value	p Value
7 days	Factor	5	6551.803	1310.361	197.9247	1.09 × 10−18
	Residuals	24	158.892	6.6205
28 days	Factor	5	7362.08	1472.41	35.13348	2.76 × 10−10
	Residuals	24	1005.823	41.9092
91 days	Factor	5	5746.858	1149.372	42.17728	4.05 × 10−11
	Residuals	24	654.0232	27.25097

).

At 7 days, mixtures with the same cement content showed no significant differences among themselves, indicating that the particle distribution coefficient had no relevant influence on compressive strength at early ages. However, among the mixtures with 15% cement (PDI 173 and PDI 204), a statistical difference was observed, with the PDI 173 mix exhibiting higher strength, possibly due to its higher cement-to-aggregate ratio, which favors early hydration.

At 28 days, mixtures with the same cement content presented statistically similar compressive strengths, further confirming that variations in the particle distribution coefficient did not significantly affect this property. Nevertheless, the influence of the packing density (PDI) on strength development becomes evident, as mixtures PDI 190 and PDI 180 achieved the highest compressive strength values within the statistical groupings. The mixture PDI 173, with its lower cement content, was unable to reach higher strength levels. Notably, mixtures with a cement content of 20%—corresponding to a cement consumption of 598 kg/m³—achieved compressive strengths of 147.03 MPa for q = 0.20 and 151.16 MPa for q = 0.25 at 91 days. These values are remarkable when compared to those reported in the literature, where cement contents close to 800 kg/m³ are typically required to achieve similar strength levels [9].

The efficiency of the UNISINOS mix design method, through the optimization of granular packing and the particle distribution coefficient (q), enables a substantial reduction in cement consumption, providing both economic advantages and lower environmental impact.

3.3. Flexural Strength Analysis

The maximum flexural tensile strength (f_p), the first-crack tensile strength (f₁), as well as the toughness and the equivalent flexural strength (RDR, 150), are presented in

Table 7. Flexural test results at 28 and 91 days for the UHPC mixtures evaluated in this study.

Mix Desing	Ages Evaluated (Days)
	Fp (MPa)		F1 (MPa)		T150		RDR, 150
	28 Days	91 Days	28 Days	91 Days	28 Days	91 Days	28 Days	91 Days
173-15%-0.20	14.43	15.98	6.51	6.97	266.47	282.79	0.1692	0.1964
180-20%-0.20	11.38	10.53	5.66	7.25	205.72	164.28	0.1444	0.1186
203-25%-0.20	13.45	17.68	9.54	9.15	246.6	304.77	0.1149	0.1802
204-15%-0.25	14.45	14.37	6.44	6.25	297.92	246.24	0.2071	0.1819
190-20%-0.25	6.32	7.69	4.68	5.75	115.33	130.39	0.112	0.111
214-25%-0.25	16.52	16.39	10.52	11.19	309.88	297.73	0.1418	0.1183

for all mixtures evaluated in this study.

The flexural test results indicate ductile behavior in all UHPC mixtures analyzed in this study, in line with observations of Hasnat and Ghafoori [36]. The addition of steel fibers enhanced the deformation capacity, such that the maximum flexural tensile strength (f_p) exceeded the first-crack tensile strength (f₁) [37,38]. Between 28 and 91 days of curing, no significant strength gains were observed, a behavior also reported by Christ [24] when evaluating different fiber volume fractions.

The results show that both the first-crack strength and the maximum flexural strength exhibited no relevant variation between 28 and 91 days, indicating that the increase in compressive strength did not directly translate to improvements in flexural strength. Similar behavior was reported by Medicis et al. [39], who observed a tendency for mixtures with higher compressive strength to exhibit reduced flexural performance. Likewise, Hasnat and Ghafoori [36] demonstrated that, for a given fiber content, there is no direct correlation between compressive and flexural strength.

It is also evident that the flexural behavior—first-crack strength, maximum flexural strength and toughness—remained nearly unchanged when comparing mixtures with the same cement content but varying only the particle distribution coefficient (q). High toughness behavior can be observed through the load–deflection curves of the tested samples, as shown in Figure 4.

The analysis of the mechanical behavior through the flexural load–deflection curves of the mixtures (Figure 4) showed that the deformation of the UHPC remained linearly elastic until the appearance of the first crack, followed by plastic deformation. The fibers act as anchoring elements, maintaining the modulus of elasticity of the matrix after the elastic limit is reached, as described by Yu et al. [40]. Thus, the fibers sustain the applied stress when cracking begins and gradually transfer it back to the matrix [41].

At 28 days, the ANOVA test (

Table 8. ANOVA for maximum flexural strength at 28 and 91 days.

		G.L	Sum of Squares	Mean Square	F-Value	p Value
28 days	Factor	5	108.1722	21.63444	15.25046	0.002341
	Residuals	6	8.51165	1.418608
91 days	Factor	5	161.0772	32.21543	40.06562	0.000156
	Residuals	6	4.8244	0.804067

) indicated a statistically significant difference among the maximum flexural strength results. According to Tukey’s post hoc test, only the mixture with PDI 190 was not statistically equivalent to the others, suggesting that variations in the coefficient q and cement content—within the levels used in this study—did not produce significant changes in maximum flexural strength. The absence of a meaningful influence when reducing the q factor from 0.25 to 0.20 on the mechanical properties suggests that similar mechanical performance can be achieved with lower binder consumption (cement and silica fume).

At 91 days, it was observed that, among all mixtures analyzed, those with PDI 180 and 190 exhibited the lowest flexural strengths, with values significantly below the group mean—an inverse trend compared to the compressive strength results. This increase in flexural strength is associated with the improvement of the ITZ (interfacial transition zone), which enhances the bond between the matrix and the steel fibers. It is noteworthy that even mixtures with higher flexural strength may display greater porosity, which can reduce compressive strength; however, the initial matrix–fiber bond effect tends to be more dominant [32,41,42,43,44].

4. Conclusions

Based on the results obtained in this study, it is possible to identify that the evaluated mix design method allows the preliminary development of UHPC mixtures with improved mechanical performance. The effectiveness of the method primarily depends on the proper selection of materials. Although the mix design approach aims to provide a practical procedure for UHPC preparation, the analysis of particle size distributions and their overlaps should be the starting point for any mixture design.

The workability of the mixtures remained within the limits reported in the literature, with flow diameters ranging from 635 to 800 mm, characterizing a self-compacting UHPC. On average, mixtures with a particle distribution coefficient (q) of 0.20 exhibited slightly higher flow than those with q = 0.25. Additionally, lower cement contents were associated with higher flow values.

Statistical analysis indicated that variations in the q factor did not have a significant effect on compressive strength at any curing age (7, 28, or 91 days). In contrast, cement content had a pronounced influence, directly affecting the compressive strength of the mixtures. At 7 days, mixtures with higher cement content exhibited greater strength due to rapid cement hydration, whereas the PDI showed no significant effect. At 28 and 91 days, the PDI became relevant: mixtures with lower PDI values (180 and 190) achieved higher compressive strengths, highlighting the impact of improved particle packing. The mix design method used in this study, by optimizing particle packing (lower PDI) and the q coefficient, allows for reduced cement consumption, promoting economic efficiency and a lower environmental footprint.

Regarding flexural strength, the q coefficient had no significant effect, while PDI exhibited an inverse trend: mixtures with higher flexural strength showed lower compressive strength. Nevertheless, most mixtures met the minimum flexural strength requirement for UHPC (>6 MPa).

In conclusion, the UNISINOS mix design method assessed in this study proved promising, particularly concerning compressive strength with lower cement consumption. The optimization of the packing deviation index (PDI), following appropriate material selection, contributes to UHPC mixtures with superior performance while maintaining adequate workability and flexural properties within typical UHPC standards.