Tribological Effects of Martian Regoliths on Stainless Steel with Natural and Composite Lip Seal and Packing Materials

Abstract

We have investigated the sealing performance of different sealants and the abrasive effect of some Martian regolith simulants (MGS-1 and JEZ-1) by pin-on-disc tests using steel and various pin material pairs. Pin materials, representing sealants, were present in the following two forms: block types, such as pure polytetrafluoroethylene (PTFE) and composite PTFE, as well as braided pins, including pure and hybrid composite PTFE. Friction resistance, wear, surface roughness, and electron microscopy analyses were used to evaluate the tribological properties. We found significant differences between the block-type and braided pin materials regarding the sealing and wear performance. The braided pins either reached stabilized coefficients of friction at a much higher level (0.55) than the block-type pins or did not stabilize at all during the longest test duration. Martian regoliths resulted in negligible wear regarding the disc, but the hybrid composite braided pins were severely worn, causing the abrasion test to be stopped before schedule.

1. Introduction

Martian regoliths present a significant challenge to the longevity and functionality of mechanical systems, including rovers, landers, and robotic equipment operating on the Martian surface. The fine, abrasive nature of the regolith, composed primarily of sharp, angular particles, leads to accelerated wear and the degradation of critical components such as sealants, joints, and moving parts [1,2].

Similarly, lunar dust exhibits extreme abrasiveness due to its jagged, unweathered particles formed by micrometeorite impacts. The lack of atmospheric weathering produces sharp edges that aggressively erode mechanical surfaces and compromise seals, leading to dust infiltration, mechanical failure, and reduced equipment lifespan. The vacuum environment and electrostatic charging of lunar dust exacerbate its adhesive and abrasive properties [3,4].

Both environments pose unique challenges. Martian dust is affected by atmospheric conditions such as dust storms and variable pressures. In contrast, lunar dust operates in a vacuum with temperature extremes and electrostatically charged particles that cling to surfaces [5]. These factors demand specialized material research to mitigate wear, focusing on durable coatings, optimized seal designs, and dust-repellent technologies. Comparing lunar and Martian abrasive effects, one can state that lunar dust particles are more angular and sharper due to the absence of atmospheric erosion. In contrast, Martian regoliths, though also abrasive, exhibit more rounded particles due to aeolian processes. As was already summarized in [6], lunar environments involve vacuum conditions with extreme temperature fluctuations and electrostatic charging, increasing dust adhesion and wear. Mars has a thin atmosphere, which leads to dust storms that transport particles at high velocities, impacting components over longer periods. Lunar dust causes more aggressive mechanical wear due to sharper particles and vacuum-enhanced dust infiltration. Martian regoliths cause erosion primarily through prolonged exposure to dust-laden winds, affecting seals and coatings over time.

Understanding these differences is critical for designing robust mechanical systems for future lunar and Martian missions, ensuring the longevity and reliability of space exploration equipment. The study by [7] assessed lunar dust’s erosive wear using NASA Glenn’s Dust Erosion Experimental Rig (DEER). The materials tested included 1045 steel, 6061-T6 aluminum, and acrylic, with JSC-1AF lunar dust simulant as the abrasive. Experimental conditions simulated erosive environments by adjusting the mass flow rate, particle velocity, nozzle diameter, and specimen distance. Differences from lunar conditions included the temperature (300 K in tests vs. 15–400 K on the Moon), pressure (760 Torr vs. 10–12 Torr), and particle velocity (105 m·s⁻¹ in tests vs. 0–2000 m·s⁻¹ on the Moon). Lunar surface chemistry was activated, while the test conditions resulted in passivation. The results showed less than 1% material loss from quarter-inch-thick specimens, but erosive wear could impact thin films or foils, emphasizing the importance of understanding lunar dust interactions for space material durability.

Another study [8] used an electromagnetic eddy current accelerator to assess lunar dust’s impact on metallic and optical materials. The specimens included AlMg₃, silica glass, acrylic, PMMA, and polycarbonate (Makrolon), with JSC-1A lunar simulant as the abrasive. The conditions included slow- and fast-moving particles under increased pressure, with about 300–400 particles per shot. The accelerator operated under a vacuum for the better simulation of lunar conditions. The results showed that silica glass had the best resistance to impact wear among the optical materials, indicating its potential for lunar protective applications.

Martian erosive effects were studied in [9] using the Particle Erosion Test Chamber. T6-aluminum with an inorganic coating served as the specimens, with structural steel as the disc. The abrasives included silica particles simulating dust (<5 μm), silt (38–53 μm), and sand (177–250 μm). The conditions controlled the particle size, velocity, impingement angle, loading, and sample translation. The results showed that the erosion was highly sensitive to the impingement angle, indicating a shear failure typical of ductile materials. The detailed study introduced the abrasive wear of Pathfinder’s wheel [10]. The Wheel Abrasion Experiment (WAE) on the Mars Pathfinder rover assessed Martian dust’s abrasiveness on metal strips attached to a wheel. A modified wheel with 15 thin film samples reflected sunlight to a sensor, measuring the dust adhesion and wear. Ground tests showed static charges of 100–300 V, but the charge dissipation points kept levels below 80 V. Dust accumulation suggested an electrostatic charging, while the photoelectric effect likely aided the slow discharge in daylight. Data showed the most wear on thin aluminum, and the least wear on thick nickel and platinum. Martian dust was found to be fine-grained with limited hardness.

The abrasive nature of Mars’ surface can significantly impact technological processes on Mars, particularly drilling technology, as previously discussed in [11]. Martian rocks hold key insights into Mars’ formation and climate history. Core drills face wear issues and risk the thermal denaturation of samples. To improve this, two new drill bits—conical straight and conical spiral junk slots—using impregnated diamonds were tested. The experiments showed that the spiral design outperformed in speed, abrasion resistance, force loading, and temperature control. Simulations on basalt confirmed the sample temperature rise.

Recent research analyses the abrasive effects of reciprocating shaft/seal contact impacted by lunar and Martian regoliths [12]. Sealants in space equipment prevent particle ingress, which is crucial for rover and robot functionality. Once a particle enters between the sealant and the shaft, it can cause severe abrasion and wear, which could be tested by several methods, as reviewed in [13].

Space industry developments have shown that the seal methods used so far react differently to dust in the lunar/Martian environment. Testing the dust resistance of seals in moving structures is crucial for various space expeditions, which is why many projects are currently addressing this issue. Several studies have investigated the failure behavior of lip steals using different finite element analyses [14,15,16] in various working conditions. The current research examined the wear performance of seal and shaft material combinations by a modified pin-on-disc test method with abrasive regolith simulants. In the pin-on-disc tests, the pin represented the sealing material, while the disc represented the shaft. Four sealant materials (PTFE, PTFE + glass fiber + MoS₂, PTFE yarn, and PTFE fiber + aramid fiber + lubricant) were tested against steel counterface using two Martian regolith simulants (MGS-1 and JEZ-1). Wear was assessed via mass loss, the pin’s head displacement measurement, and surface topography.

2. Materials and Methods

2.1. Materials

A series of modified pin-on-disc tests were conducted for 2, 6, 15, and 30 min to evaluate the abrasive impact of Martian regolith simulants. The disc (Ø100 × 12 mm) used in the tests was made from 316 L (ASTM) stainless steel, chosen for its austenitic structure, which offers excellent mechanical performance at low temperatures.

For the pin materials, two types of sealing solutions—lip seals and packing materials—were selected based on prior research [4] and recommendations from the European Space Agency (Noordwijk, The Netherland). The selected lip seals commonly used in standard-grade tribosystems consisted of natural PTFE (labeled Ln) and a PTFE composite with 15% glass fiber and 5% MoS₂ (labeled Lc). Additionally, the following two high-performance packing materials were included for harsh, dry sliding conditions: PTFE yarn with dry impregnation (Pn) and a PTFE hybrid composite combining PTFE fibers, solid lubricants, and aramid fibers (Pc) to enhance the edge durability and maintain the dimensional stability under press-fit conditions.

The abrasive media used in the tests were two Martian regolith simulants—MGS-1 (Martian Global Simulant) and JEZ-1 (Jezero Delta Simulant)—sourced from Exolith Lab in Orlando, FL, USA. The key properties of these simulants are summarized in

Table 1. Main characteristics of the regolith simulants.

Regolith Simulants	Mineral Phases	MGS-1	JEZ-1
		Martian Global	Jezero Delta Simulant
Phase composition	Anorthosite	27.1%	16.0%
	Glass-rich basalt	22.9 %	13.5 %
	Pyroxene	20.3%	12.0%
	Olivine	13.7%	32.0%
	Mg-sulphate	4.0%	2.4%
	Ferrihydrite	3.5%	2.1%
	Hydrated silica	3.0%	1.8%
	Magnetite	1.9%	1.1%
	Anhydrite	1.7%	1.0%
	Fe-carbonate 1.4%	1.4%	-
	Mg-carbonate	-	11.0
	Smectite	-	6.0
Physical properties	Bulk density (g/cm3):	1.29	1.54
	Median particle size (μm):	60	60
	Particle size range (μm):	>0.04–600	<0.04–500
Reference		[17]	[18]

.

2.2. Test and Analysis Methods

Tribological tests were conducted using a modified pin-on-disc setup, with measurements performed on coupons. The tribo-testing machine was a customized pin-on-disc device, which was illustrated in detail in a previous paper [19]. The testing approach followed a standardized testing protocol in line with [20]. A sample preparation process was implemented to standardize the clean surfaces of the shaft material and sealings. An ultrasound bath (Proclean 10.0 MS, Expondo Polska, Zielona Góra, Poland) and 1% EM-300 Metallreiniger (EMAG AG, Mörfelden-Walldorf, Germany) cleaning liquid were used to remove any milling and machining chips and grease. Followed by rinsing with distilled water and drying, isopropyl alcohol was applied to remove traces of oil or fingerprints from the shaft material surface. A 24 h treatment in a DRY30EA cabinet (Changsu Catec Electronic Inc., Suzhou, China) at 34 °C was applied to remove traces of isopropyl alcohol. Preliminarily, the necessary volume of regolith was determined to cover the whole disc area with a thickness of 6 mm. During rotation, a dust deflector flap was used to guide, collect, and concentrate the dust in the same orbit above the wear track so to ensure that the disc surface was uniformly coated with regolith powder. This setup allowed for the analysis of the intricate interactions that occur when flat-surfaced simulated materials slide against each other.

We investigated the mechanisms of friction and three-body abrasion, surface abrasion, contact zone deformation, and surface changes on a micro-geometric level. To analyze the embedding capacity, abrasive impact, adhesion, and grain dynamics of the simulants, the SEM (Zeiss EVO 40, Carl Zeiss SMT Ltd., Cambridge, UK) and EDX (JEOL JSM-IT700HR, JEOL Ltd., Tokyo, Japan) techniques were used.

In the pin-on-disc tests, the pin representing the sealing material was a piece of cube (8 × 8 × 8 mm; Figure 1) that was pressed against the rotating (n = 36.6 rpm) disc representing the shaft. The sliding speed of the pin rotating along a circle of 25 mm radius track was 0.1 m·s⁻¹. The rotating disc was made of the same material as the modeled shaft. Further details and a figure of the system can be found in [21]. To maintain the abrasive conditions, a boundary ring was attached to the edge of the disc, allowing the surface to be covered with abrasive particles approximately 3 mm thick. As the disc rotates, diverting plates continuously guide the regolith particles back onto the sliding track. These particles become lodged between the shaft material (disc) and the sealant surface, creating an abrasive effect similar to that experienced in real rotating shaft/seal contacts. During the experiment, changes in the abrasive friction resistance—analogous to torque fluctuations in a rotating shaft—and wear involving layer transformation on the seal and disc surfaces were monitored in real time. The friction force (in N) was measured using strain gauges configured in a full-bridge setup, while the vertical displacement of the pin’s holder head (referred to as wear) was recorded in millimeters using an inductive transmitter. In this case, “wear” includes the deformation as well as the 3rd-body effects. Data acquisition was conducted online using a Spider 8 system.

A 3D optical microscope (Keyence VR 5200, Osaka, Japan) was employed to analyze the surface’s two- and three-dimensional topography parameters. The 2D surface topography data were used to determine the degree of penetration (Dp), a parameter that characterizes the micro-mechanism of the wear, as detailed in [22]. The Dp provides insight into how abrasive particles interact with the material’s surface, influencing the type of wear. The Dp was calculated as Rz/0.5Rsm, where Rz is the ten-point height [µm] and Rsm [µm] is the mean spacing at the mean line of the surface. It was found that, regardless of the attack angle, a critical Dp value of approximately 0.2 could be used to differentiate between the following two primary wear mechanisms: micro-ploughing and micro-cutting. When the Dp is below this threshold, micro-ploughing occurs, where the abrasive particles deform the surface without the significant removal of material. Conversely, when the Dp exceeds this value, micro-cutting dominates, leading to the removal of material as chips. All of the modes of abrasive wear can generate grooves on the material’s surface due to particle interaction and the plastic flow of the material. This process forms characteristic ridges along the edges of the grooves, highlighting the material’s response to abrasive forces [21].

2.3. Regression Model

Linear and multiple linear regression models were developed using IBM SPSS 29 software for the statistical analyses. Only the 30-min runs were considered, as they exhibited nearly parallel behavior compared to other experiments of varying lengths. Additionally, running-in phases were excluded from the analysis. Notably, the running-in phase’s duration was relatively short compared to the overall test duration and, more importantly, to the expected operational lifespan of the machinery made from materials like stainless steel. Therefore, its exclusion did not introduce any significant bias into the analysis. A moving average with a window length of 500 data points (equivalent to 5 m) was applied to smooth the measured on-line data and enhance the trend visibility. Smaller window lengths, such as 200, produce similar results with slightly reduced fitting precision. However, excessively narrow windows, such as 50 (a smaller scale), lead to inadequate data smoothing, making it difficult to identify trends clearly.

In the models, the displacement of the pin holder (hereinafter referred to as “wear”) and the coefficient of friction were considered dependent variables. The independent variables included the sliding distance (d), material properties of the seal material cut pieces (the Shore D hardness, elongation at break, and compressive strength at a 1% deformation), and the relative mass fraction of regoliths with different particle sizes (PRx, where x denotes a specific particle size range; see details in the relevant sections of the results).

In a regression model, the R² value represents the goodness-of-fit, indicating how well the model aligns with the observed data. In the case of multiple linear regression, potential cross-effects among the independent variables make it difficult to isolate the pure effect of any single variable. However, despite this limitation, the relative magnitude of each variable’s influence can be estimated by examining the absolute values of the standardized (beta) coefficients for significant independent variables. A higher absolute beta coefficient indicates a stronger effect size.

3. Results and Discussion

3.1. On-Line Friction

3.1.1. Block-Type (Ln, Lc) Pins

Figure 2 and Figure 3 show the effect of the MGS-1 and JEZ-1 regolith simulants on the frictional performance of the various stainless steel disc/sealant tribopairs. For each tribopair tested, the abrasive pin-on-disc tests were carried out for different periods (2/6/15/30 min with a constant speed of 0.1 m·s⁻¹ and a load of 0.2 MPa). Only the plots corresponding to the most extended test duration for each simulant powder are presented here, while the graphs of the particular tribopairs for restricted time periods are shown in the Supplementary Materials (Figures S1–S8). The friction coefficient is, in fact, a calculated abrasive friction number characteristic of an open three-body abrasive sliding mechanism. The graphs are plotted using a moving average over a period of 60 data points. The curves are different not only for different steel and sealant tribopairs but also for the different regoliths.

The natural PTFE pins (marked as “Ln” hereinafter—the natural material of the lip seal) exhibited different frictional behaviors depending on the type of the regolith. In the case of the MGS-1 simulant, the running-in phase was around 10 m, reaching a maximum friction coefficient of 0.35–0.4. After 35–45 m, the particles within the formed layer began to shear and roll, temporarily reducing the resistance without leading to stabilization. The layer adhering to the metal surface (Figure 4a) appeared more uneven, with increased particle migration and jamming, which raised the frictional resistance during the stable stage of the process, ultimately stabilizing at approximately 0.35–0.4. The impact of sliding–rolling grain marks was also evident on the steel surface, manifesting as abrasive scratches and grain embedment. In contrast, the JEZ-1 simulant exhibited a fundamentally different behavior. Transient processes were observed even during the running-in phase, likely due to dust infiltration, temporary adhesion, and modifications in the initial adhesion. After this phase, relatively few particles remained in the contact zone, as clearly shown in the surface photographs (Figure 4b). During the stable sliding phase, frictional resistance followed a linear decreasing trend, with the abrasive friction coefficient declining to 0.2.

Out of the tested pins, only the composite PTFE pin (Lc: PTFE/15%GF + 5%MoS2—material of composite lip seal) showed a similar behavior for both regoliths. The running-in section can be clearly identified, but, compared to the natural PTFE test specimens, the phenomenon occurred in a wider friction path length range (8–15 m) (Figure 2 and Figure 3), which shows the uncertainty of the contact zone behavior. In the moments after the start, the pure and clean contact sliding turned to abrasive friction. When the particles entered the contact zone, the friction jumped to around 0.4–0.45. After that, a temporary dropping trend came, followed by a moderate but continuous increase or a steady-state of the friction for the MGS-1 and JEZ-1 regoliths, respectively. The final stabilization of the friction dynamics was expected around 0.45 for MGS-1 and 0.4 for JEZ-1 simulant, which is considered a high friction resistance.

3.1.2. Braided (Pn, Pc) Pins

For the natural PTFE packing pin (Pn—the cut piece of natural PTFE packing), the running-in stage was obtained at about 10 m, at which point the friction process stabilized at a friction coefficient slightly above 0.5 for the MGS-1 regolith. For the JEZ-1 regolith, friction stabilization occurred only after 50–80 m at a value of 0.55. Subsequently, oscillations emerged due to transformations in the contact zone, where the structure between the fibers and its grain-embedding capability created local transient peaks and frictional instability. By the end of the test period, the packing structure had completely disintegrated.

When comparing the system’s results with the friction curves of the natural PTFE block pin (Ln), it is evident that friction stabilized at a higher level for the packing braid than for the block material, but with greater uniformity (exhibiting smaller fluctuations and fewer local instabilities). In both cases, the frictional resistance eventually stabilized, though with differing degrees of variation and absolute values.

In the case of the composite packing pin (Pc: PTFE/aramid/solid lubricant-cut piece of the hybrid composite PTFE-based packing), the running-in section was barely recognizable (1–2 m), and the frictional resistance continuously increased in accordance with the phenomenon associated with the ingress of the regolith dust and the sealing of the secondary braided assembly. After covering a sliding path length of a few tens of meters, frictional instability arose, accompanied by the disintegration of the braided secondary structure. Due to the premature disintegration, we had to stop the measurement after 90 m and 120 m for the MGS-1 and JEZ-1 regoliths, respectively.

3.2. On-Line Wear

3.2.1. On-Line Wear of Block-Type (Ln, Lc) Pins

Figure 5 and Figure 6 show the vertical movement of the pin holder, indicating wear (including deformation, particle motion, and thermal expansion) with respect to the distance traveled. Wear for shorter test durations is shown in Figures S9–S16. At first sight, the wearing behavior of the block pins (Ln and Lc) seems quite similar, having almost negligible wear regardless of the type of the Martian regolith. However, there were some minor but definite differences. The MGS-1 regolith entering into the abraded region moved up the pin fixing unit, which resulted in an apparent negative wear for both the Ln and Lc pins in the sliding range of 30–180 and 10–80 m, respectively. During this period, in accordance with the friction curve, the layer formation and particle accumulation can cause a local fluctuation of the wear curve. While in the test system with the Ln pin, the zero level of the specimen holder is just reached at the end of the test run, and the cyclic migration of the third-body can be observed for the Lc pin, suggesting chaotic third-body dynamics in the contact zone for the longer periodic time. With respect to the JEZ-1 regolith, the photos of the steel (Figure 4) and PTFE surfaces (Figure 7a,b) clearly show that a significantly lower number of particles remain in the contact zone. Consequently, no measurable change in the wear was observed in the system. Wear on the PTFE specimen remained minimal, with reduced three-body effects, ensuring stable wear curves and minimal measurement deviations.

3.2.2. On-Line Wear of Braided (Pn, Pc) Pins

The wear behavior of the system with the Pn packing was affected by the deterioration and deformation of the pin structure (Figure 7c,d), which is a steady, quasi-linear process. When using the MGS-1 simulant, at the beginning of sliding, before the dust entered the contact zone, the deformation of the pin showed positive wear during the running-in. It rapidly turned to a decrease, reaching a negative wear due to the incoming dust, which raised the pin holder head of the test specimen. After this, a steadily increasing, stable wear curve was formed due to additional pin deformations and dust layer migration. The JEZ-1 simulant induced a different wear trend: the positive wear resulting from the initial deformation started to stagnate after the running-in. Then, it decreased due to the impact of the particles entering the rubbing zone, which raised the pin holder head back, although it did not reach the negative wear range. Further, the continuous deformation and the shearing of the adhered regolith layer, the three-body abrasion mechanism, moved the process in the direction of positive wear again.

System wear behavior with the Pc braided packing was also greatly affected by the tendency of the pin for disintegration. In case of the JEZ-1 regolith, at the beginning of the measurement, the braided structure began to deform and disintegrate, resulting in a “wear” with a positive sign. The lubricant between the fibers boosted this mechanism. As a result of the regolith dust trapped between the fibers, preventing the dynamic particles from moving, and the thick layer of regolith that accumulated, the wear slowly turned to the negative direction. The photos taken of the pin surfaces (Figure 7c,d) clearly illustrate the crushing of the braided structure, the presence of a thick adhered layer of dust resulting from the combined presence of lubricant and regolith dust, and the dynamics of the surface zone. In the case of the MGS-1 regolith, the adhesion started randomly on the surface of the crushing pin, which suddenly brought the value of “deformation and wear” into the negative range. With the disintegration of the braided pin fibers, the regolith accumulating on the pin surface caused by the impregnated wax and solid lubricant made further measurement meaningless.

3.3. Surface Analyses

Light microscope images of steel discs after the pin-on-disc tests (Figure 4) confirmed the surface phenomena seen in the friction and wear curves, including adhesion, particle dynamics, and layer formation. The wear track on the steel surface appeared polished, with noticeable topographical transformation due to abrasion. The reduction in the roughness parameters (

Table 2. Surface roughness parameters of the steel disc after the 30 min test period.

Pin/Simulant	Sa		Sz		Sq		Ssk		Sku
	Before	After	Before	After	Before	After	Before	After	Before	After
Ln/MGS-1	1.02	0.70	171.64	14.13	1.85	0.89	3.65	−0.32	179.34	4.31
Ln/JEZ-1	1.02	0.71	166.87	9.32	1.92	0.88	3.82	0.09	185.56	2.80
Lc/MGS-1	1.01	0.68	171.11	15.30	1.87	0.85	0.18	−0.05	178.54	3.32
Lc/JEZ-1	1.04	0.71	185.15	14.66	2.84	0.90	3.11	0.01	194.42	3.06
Pn/MGS-1	0.92	0.74	127.11	9.26	1.41	0.79	3.55	−0.05	149.75	2.91
Pn/JEZ-1	0.98	0.66	126.22	9.01	1.28	0.82	3.02	0.02	141.56	2.70
Pc/MGS-1	1.08	0.79	162.77	86.50	1.90	1.00	2.71	−0.01	170.04	41.30
Pc/JEZ-1	0.99	0.70	133.12	42.44	1.55	0.90	3.01	0.01	178.08	52.08

) suggests a significant polishing effect across all of the tested systems, smoothing out the higher peaks and balancing the distribution of peaks and valleys. However, the slightly smaller decrease in the roughness parameters for the systems with the braided sealant suggests that the abrasive effect of the dust was more intense for those systems with block pins. The SEM images of the steel disc after the longest abrasion tests with the particular sealants also confirm the greater abrasive effect of Martian regoliths the on steel/block-type sealant tribopairs. Increasing surface damage and uneven wear marks reflect the cutting effect of the regolith particles stuck to the PTFE pin (Figure 4). Based on multi-line groove analyses, the calculated Dp values were always less than 0.1 (

Table 3. Degree of penetration values for the steel disc after different test periods.

Dp, (Rz/0.5Rsm)	Martian Simulant	Test Period (min)
		2	6	15	30
Ln pin	MGS-1	0.046	0.039	0.047	0.041
	JEZ-1	0.058	0.056	0.050	0.048
Lc pin	MGS-1	0.061	0.055	0.042	0.043
	JEZ-1	0.058	0.061	0.051	0.059
Pn pin	MGS-1	0.042	0.041	0.039	0.038
	JEZ-1	0.038	0.036	0.044	0.041
Pc pin	MGS-1	0.050	0.053	0.034	0.028
	JEZ-1	0.025	0.025	0.062	0.057

), which means that the dominant surface transformation of the steel surface was micro-ploughing.

The amount of regolith sticking to the Ss disc was considerably smaller (Figure 4a,b) for the braided pins (Pn and Pc). The surface remained almost unaffected by the abrasion test. The regoliths bonded mainly to the braided pin’s surface and valleys (Figure 7c,d), a process accelerated by the wax and solid lubricants. Its abrasive impact on the steel disc was minimal due to the pin’s flexibility. In parallel, pin fragments that came off the braided structure were embedded in the scratches of the disc (Figure 8, SEM).

Similar to those cases for the block pin materials, the “degree of penetration” < 0.1 values (Table 3) indicate massive micro-ploughing abrasion, with plastic deformation and ridge formation. However, the polymer-adhered layers and embedded regolith particles can fill the wear grooves, smoothing the surface and lowering the degree of penetration (Dp).

The pin material parts are the most susceptible to wear and/or deformation in the tribological systems investigated because of their moderate-to-low hardness values. While PTFE was the primary component of all the tested pin materials, their wearing behavior showed great variations depending on the secondary structure. The main difference between the block pin and packing materials is that the latter are prone to deformation and disintegration after prolonged use. Images (Figure 7) of the woven pins after 15 min of testing show deformed, torn packing with noticeably accumulated abrasive dust embedded in the surface. The surface and geometry indicate that the packing is no longer functioning properly, making the surface roughness analysis irrelevant due to the significant topographical distortion. The damage is also shown in 3D images (Figure 9). The original braided (secondary) structure is typically disappeared and fully covered with regolith.

Block-type pin materials were much less affected by the regolith particles than the braided packing pieces. The regolith particles covered the PTFE over the whole contact area, but some of the accumulated bigger particles caused deformation and a cut surface (Figure 4c). An increase in the test time resulted in more uniform abrasion groove patterns. Embedded particles were more clearly identified with an increase in the test runs. The dominant surface transformation of the counterface was mixed in all of the runs. The Dp values (

Table 4. Degree of penetration values for the pins after different test periods.

Dp, (Rz/0.5Rsm)	Martian Simulant	Test Period (min)
		2	6	15	30
Ln pin	MGS-1	0.214	0.223	0.229	0.205
	JEZ-1	0.142	0.128	0.165	0.158
Lc pin	MGS-1	0.179	0.242	0.268	0.305
	JEZ-1	0.124	0.115	0.298	0.165

) indicate a transformation/mixed mechanism of abrasive wear. The surface experienced plastic deformation primarily through micro-ploughing, resulting in distorted surfaces and the formation of ridges. This process occurred as the material interacted with hard soil particles during sliding, causing the surface to be gradually reshaped. Simultaneously, material was removed via micro-cutting, where the soil particles acted as abrasives, effectively scraping and dislodging material from the surface. As the sliding distance increased, the extent of micro-cutting became more pronounced, leading to further material removal. This increase in the sliding distance was reflected in the Dp values, which indicated a higher level of micro-cutting and material loss over time.

3.4. Regression Model Results

Multiple linear regression analysis was carried out to identify the most important parameters that affected the coefficients of friction and wear. The included independent variables were the distance (d) traveled by the pin and the relative ratios of the regoliths with different particle size ranges (PRx). These relative ratios were considered constant during each test.

As the first step in each regression analysis case, the F test was carried out to see whether the considered linear model was relevant or not, and whether they differed from a constant model. The F tests were significant for all but one case, p < 0.001, and for an exceptional case, p < 0.048. Furthermore, the respective analysis of variance tests were highly significant (p < 0.001), except in the same singular case (p < 0.048), which also indicated that the variance explained by the linear models was relevant.

In the case when the distance was the only independent variable, the goodness-of-fit varied between 0.2 and 0.97 among the investigated tribopairs and regoliths. The explanation for the few seemingly not-so-good results is that, although the linear model was relevant, the slopes of those fitted lines were nearly zero, that is, the fitted function was almost constant, and there was nearly no “wear” or frictional effect in these cases.

Using the multiple linear regression model, the most influential regolith size intervals (

Table 5. Model equations for the wear and friction coefficient for the tested pin materials.

Pin	Dominant Regolith Size *	Model Equation	R2	$\mathbf{Standardized}\mathit{\beta }$ Coefficient of d	$\mathbf{Standardized}\mathit{\beta }$ Coefficient of PRx
Ln	PR400	wear = 2.706 + 0.0003d − 100.396 PR400 CoF = −3.576 + 0.0001d − 139.356 PR400	0.922 0.849	0.255 0.085	−0.962 0.906
Lc	PR400 PR1000	wear = 0.209 + 0.0001d − 8.044 PR400 CoF = 0.441 − 0.00006d − 2.361 PR1000	0.4 0.569	0.514 −0.12	−0.402 −0.752
Pn	PR200	wear = −0.884 + 0.002d + 4.276 PR200 CoF = 0.373 + 0.00008d − 0.888 PR200	0.517 0.74	0.709 0.142	−0.202 0.835
Pc	PR125	wear = 1.501 − 0.001d − 3.818 PR125 CoF = 0.37 + 0.001d	0.555 0.926	−0.701 0.663	−0.088 -

* PR₁₂₅: 125–200 µm; PR₂₀₀: 200–250 µm; PR₄₀₀: 400–500 µm; PR₁₀₀₀: >1000 µm.

) were identified, showing that they were the same for both the wear and friction for the Ln and Pn sealants, while, for Lc and Pc, they were different. Interestingly, in the case of Pc, the regolith particle size did not affect the coefficient of friction. Our results point out that the behavior of Ss was influenced differently depending on the pin type. For Pc and Pn, smaller regolith fractions had the most significant impact, whereas, for Ln and Lc, the larger fractions played a more significant role. Although, the goodness-of-fit was not so high in some of the cases, our models still provided information on the trends of the wear and friction.

4. Conclusions

In this work, we investigated the abrasive impact of Martian regoliths (MGS-1 and JEZ-1) on stainless steel/PTFE-based sealants, investigating four different sealing materials, including the block and braided type, as well as homogeneous and composite PTFE sealants. The major findings were as follows:

• The coefficient of friction of the tribosystem stabilized at a typically high value of 0.3 to 0.6 for all systems, except for the hybrid composite braided packing sealant, when the frictional resistance continuously increased due to the ingress of the regolith and the fractions of the detached sealing of the secondary braided assembly.
• The abrasive effect of the Martian regolith depended primarily on the type of sealant and much less on the type of regolith simulant. In the case of the block-type sealants, the disc part, while, in the case of the braided sealants, the sealant part of the tribopair was more affected by the abrasive effect of the regolith. For the block sealants, the regoliths showed only small particle incorporation at the conclusion of the run-in period, which resulted in a low-slope, nearly horizontal, wear curve with no measurable significant wear. The braided packing materials were prone to disintegration during the abrasive test. Thus, it is not possible to speak about the real wear curves. Apparent wear occurred due to the dynamic effects of the regolith powders in the contact zone and the disintegration of the pin specimens.
• In designing any tribosystems for the Martian environment, it must be kept in mind that block-type sealants perform better as compared to any braided packings, while there is negligible difference between the different block-type pins.