Development of fused metrology methods for the analysis of hip implant tribology

In order to advance the study of wear phenomena involved in total hip implants a fused metrology system has been designed and constructed. The novel fixture system has been designed and built to facilitate large area surface measurement for hip implant bearing wear. The system allows coordinate measurement machine (CMM) wear map data to be used for precise positional targeting of areal surface metrology using high spatial resolution optical interferometry. The complete measurement process allows Coordinate Measuring Machine measurement and surface metrology datasets to be ‘fused’ thus facilitating carefully positioned wear scar analysis. The fixture utilises two digital rotary stages, in a gimble configuration, to precisely position the bearing component during CMM measurement and surface metrology. To test the effectiveness of the fixture system a cohort of explanted large metal-on-metal (MoM) femoral heads were assessed. Application of the fixture system allowed a set of grouped surface measurements were taken within the wear area, the wear area boundary region, and at unworn locations across the femoral heads. Additionally, a series of stitched surface measurements are taken through the entire wear area and combined into a single surface measurement. The ‘fusion’ method, allowed areas of roughening (or smoothing) to be estimated and overlayed on the corresponding CMM wear map. The developed fixture system allows for better understanding of hip implant performance. Combined with areal surface metrology parameters such as areal average surface roughness Sa, developed surface area ration Sdr and areal surface skewness, Ssk the system could be utilised understand the wear mechanisms for both explanted, in-vitro and in-vivo wear testing and also detailed quality inspection of newly manufactured components. The significance of the system deployment is that wear location and detailed mechanism can be studied simultaneously, thus delivering understanding of typical wear mechanisms and hence the fixture becomes a tool for developing increased implant life through understanding surface interactions.


Introduction
According to National Joint Registry (NJR) data, hip and knee replacement surgeries account for approximately 98% of all joint replacement surgeries in the UK. In England and Wales there are approximately 160,000 total hip and knee replacement procedures performed each year. The National Joint Registry UK (NJR) data points out that osteoarthritis was given as a documented indication in 91.9% of primary hip replacement surgeries [1]. The demand for joint replacement is increasing with the percent increase in total annual United States primary total hip replacement is projected to be a 284% rise between 2019 to 2040 [2].
Total hip replacements are made up of several key components as shown in figure 1(i) an acetabular cup (C) made of ultra high molecular weight polyethylene, ceramic or metal and in combination with a secondary liner (B), is implanted/cemented into the acetabular hip bone (A). (ii) Running against the acetabular surface is a femoral head component (D) either metallic or ceramic, transmitting the normal, bending and torque loading to the acetabular (iii) The head is attached or mounted onto a femoral stem (E) which is fixed via Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. polymethacrylate cement into the femur or is a cementless design for facilitating osteo integration (F).
Ideally an implant will last as long as possible, restoring patient's joint mobility and relieving their pain. Currently patients can expect a hip replacement to last 15-20 years in around 75% of cases and last 25 years in 58% of cases, before revision may be needed. Periprosthetic bone resorption resulting in loosening of the implant, otherwise known as aseptic loosening, is regarded as the most common cause of revision worldwide [3,4]. Aseptic loosening and other related failure factors (wear, lysis, soft tissue reaction) is accelerated in the presence of wear debris and account for the majority of causes for revision surgery in the UK, [1,3]. The majority of wear debris is generated primarily at the head/acetabular cup bearing interface, but wear debris can also be produced at nonarticulating surfaces [4]. It is therefore a critical aim for manufacturers to keep wear rates of implants as low as possible and that during wear simulation and explanation manufacturers need a suitable means to quantify wear phenomena.
Due to the prevalence of premature implant failure induced by wear particles, a common focus of hip implant study is the wear performance of the implant [1][2][3][4][5]. Coordinate measurement machines (CMM) are commonly used for geometrically located volumetric wear measurements [6], allowing a wear map of the bearing surface to be created as well as an assessment of the overall wear volume [7]. Additionally, to understand changes in implant surface roughness resulting from wear mechanism changes and tribological interactions on a surface scale, optical surface measurements are utilized [8]. From component size mismatch occurrences, to general use, the surface topography of the bearing components has long been shown to affect the overall wear rates of the bearing through both simulator and retrieval studies [7].
It has been well established that the surface topography of the hip implant bearing changes within wear areas [9][10][11][12][13][14][15][16][17][18][19][20][21]. Therefore, when surface measurements are taken on worn bearing surfaces, they need to be positionally targeted both within and outside the wear area to give a comprehensive analysis of the entire component and to highlight tribological interactions. Currently precision positioning of surface topography surface measurement is not possible. This is the motivating factor of the present study.
For detailed assessment of the wear phenomena occurring at the bearing interface, theoretically, the CMM wear map can be used as a positional reference for surface measurements. This should allow surface measurements to be targeted within the wear area and the surface measurement location tracked with reference to the wear map. The positional targeting of surface measurements is achieved by manually and by visually manoeuvring the hip implant component with the aid of a suitable fixture (part holding device). A common theme among scientific studies in this area, is a lack of consideration of the positional repeatability of the fixture in relation to the perceived wear area and thus detailed studies of wear mechanisms in relation to position is virtually impossible.
Most manufacturing companies engaged in orthopaedic manufacture have access to high quality CMMs and surface metrology instruments. In fact, these are required for compliance with company and international quality standards [22,23]. However, in general they do not have the ability to combine outputs of these systems in a fused data context to allow them to investigate tribological/wear behaviour. Non-contact measurement can be particularly useful in orthopaedic surface measurement as there is no possibility of implants being damaged by a mechanical probe. This may be a serious issue when measuring orthopaedic materials such as ultra-high molecular weight polyethylene (UHMWPE). A specialist instrument used in the orthopaedics industry capable of such measurement is the REDLUX optical CMM (https://redlux. net). The RedLux CMM scan is designed to work in a spiral pattern and is also set to measure up to 90°elevation. The centre of the spiral originates at the pole of the component and spirals outwards towards the rim of the component. The system is relatively fast and is capable of measuring wear volume and position having comparable results with conventional high positional accuracy contact CMM's. However due to the RedLux's lack of spatial resolution it is not capable of measuring surface topography [24].
The aim of the present study is to develop an advanced computer controlled fixture system specifically designed for high lateral resolution (sub 50 um), wear mapping across large areas of hip implant surfaces. The system uses CMM data to 'guide' the fixture positioning to deliver a fused metrology approach.
The fixture system along with associated software facilitates stitching of surface measurements through wear areas, and at specified surface locations. This capability is a novel step in analysis of worn hip implant bearing components. To test the effectiveness of this fixture system, a cohort of explanted MoM femoral heads has been measured using this newly developed fixture capability for CMM guided surface metrology. The functioning system fixture system will improve future hip implant surface measurement studies, ensuring the most critical surface topography data is captured efficiently.

Custom fixture system and operation
The aims of the fixture system is to use a coordinate measurement machine (CMM) to measure an entire bearing surface on a total hip implant. This CMM data is then analysed using custom software to deliver the volume and position of wear areas on the surface of the implant. Using the CMM data the fixture can then select and position areas on the hip implant surface, under the lens of a surface interferometer for detailed topography measurement.
The designed and constructed fixture, figure 2(a) utilises two rotary stages (azimuth/elevation), in a gimble configuration, to position the femoral bearing component, allowing for access to the entire bearing surface. The frame of the fixture is custom built to be small, 271 ×155 × 93 mm (LWH), allowing it to be compatible with the majority of optical interferometry surface profilers. High precision digital rotary stages are used, Standa 8MR174-11, resolution: 0.015°, ensuring sufficient precision for co-registration of CMM and surface datasets.
The measurement proceeds via initial CMM measurement of the hip component with the fixture optimally aligned on the CMM. When the CMM measurement is complete, the component is not removed from the fixture, instead the entire fixture is moved to the surface metrology instrument with the component still attached securely in place. To reduce azimuth alignment error between the CMM Figure 2. CAD image of custom fixture system. measurement and the surface measurement coalignment between the measurement planes of the two systems is critical.
As the surface measurement instrument base will not be perfectly square with the camera sensor on placement of the fixture on the optical interferometer, the camera sensor system of the interferometer is utilised for alignment purposes. This results in a further reduction in the azimuth alignment error. To achieve this, an alignment slide mounted kinematically with the aid of a magnet on top of the fixture is used. The slider is laser engraved with orthogonal fiducial markings which can be aligned using the rotational stage, to the interferometer camera axes. Finally, to allow for different femoral head sizes a manual z adjustment (height figure 2) is mounted below the azimuth rotary stage and a mounted laser distance measurement sensor is added. The laser is mounted so it aligns with the elevation axis, this allows the user set the height of the axis of elevation when the femoral head size changes, figure 3.
The fixture system was tested extensively to determine its positional error. This was calculated as ±0.04021°(mean error value plus two standard deviations for 95% coverage). This level of precision allows for efficient stitching (20% overlap) of surface measurements Through the use of the digitally controlled rotary stages, custom software was developed for advanced data fusion control of the fixture, figure 4. The software firstly creates a wear map from CMM measurement data. The user can then select their desired surface measurement locations based on that wear map (black dots), either with spherical coordinate input or via mouse cursor selection. The fixture is then driven to the selected locations quickly and with suitable precision via the software and fixture actuation.

Cohort
Ten large explanted Metal-on-Metal (MoM) femoral heads (Zimmer Durom's) were analysed using the fused metrology approach, with a bearing size ranging from 42 to 54 mm diameter. Upon visual inspection no obvious wear areas was noted, however, small scratches were present on the surface, their location and direction appeared random. For the purposes of this methodology-based study, clinical data was known but not considered relevant.

CMM measurement
The femoral heads were initially measured using a Zeiss Prismo CMM, this instrument has a length  measurement error of 0.9 + L/350 um, and a scanning error of 1.3 um [25]. A polar measurement strategy was utilised, with scan lines from the rim to the pole of the femoral head in a circular pattern covering up to 90°elevation. A total of 400 equally spaced scan lines are used with 394 points per line, resulting in a total of 157,600 probing points. This strategy results in an estimated measurement uncertainty of +/− 1.859 mm 3 according to the method previously reported by Bills et al [7].
The analysis of CMM measurement data was completed using the software CATIA [26]. According to Bills et al [7], a reconstruction of the original surface is created from the measurement data, if the femoral head exhibits a wear area, this is excluded from the reconstruction data. A deviation analysis between the reconstructed original surface and the measurement data allows for a wear map of each femoral head to be created and the relevant parameters such as femoral head radius wear depth and wear volume can be calculated [7].

Surface measurement
To demonstrate the utility of this approach the following procedure was adopted post CMM measurement. Surface measurements were taken with a Taylor Hobson Talysurf CCI [27], using a 20× lens, (NA = 0.4) resulting in an effective measurement field of view of 918 × 918 um with a sampling spacing of 1.8 um (512 × 512 measurement points). The custom fixture system was utilised to manoeuvre and position the femoral heads before each surface measurement facilitating the assessment of specified positions on the head, grouped area measurements and measurement stripes. All surface measurements were analysed using the software Surfstand [28], a gaussian SF filter was applied, (S filter nesting index = 0.008 mm and an F operator giving a bandwidth of 100:1, following ISO recommendations) [29], and the Sa, Sdr, and Ssk values were recorded. Surface measurements were stitched together using the Surfstand software.
Sa; mean areal surface roughness, measures the roughening that may occur within a wear scar.
Sdr; Developed interfacial area ratio, is defined as the ratio of the increment of the 'real' interfacial area of measured surface within the defined area over the measurement definition area. Sdr is useful in describing the roughening of a surface as a result of wear.
Ssk quantifies the asymmetry of the areal surface heights about the mean measurement plane. The sign of Ssk indicates the predominance of peaks (i.e. Ssk > 0) or valley structures (Ssk < 0) comprising the surface. It is useful in understand wear phenomena.
The surface measurements were split into three stages:

Unworn head measurements
Each femoral head was measured at 25 equally spaced locations, 5 of these were taken at the rim. If a wear area was deemed present according to the CMM measurement, this area was avoided. This measurement set allows a baseline for the 'unworn surface' of each femoral head to be established.

Worn head measurements
For each of the worn femoral heads (where the wear area was identified from CMM measurement) the following locations were assessed: Grouped Measurements

CMM measurements
From the ten femoral heads assessed, only two showed clear wear areas, as delineated by the CMM measurement data, femoral heads A and B. The size and shape of these wear areas can be seen on the overlay plots, figure 5, designated by the darkest areas using greyscale contours. The wear volume on both of these femoral heads is relatively low, table 1, this to be expected according to a similar MoM retrieval study [30]. Of the remaining eight femoral heads, the level of wear on six was undetectable. Undetectable wear highlights two limitations 1. no wear was present, or the wear is below the spatial resolution limit if the Ziess CMM used (Zeiss Prismo 5), 0.9 + L/350 um). From the CMM measurement, the level of wear on the final two was within the uncertainty limit of the CMM. Heads A and B, with measurable wear, are used for detailed surface analysis using the custom fixture system.

Unworn surfaces
For the 'unworn' set of measurements across the whole cohort of 10, the median Sa value for each femoral head ranged from 3.55 nm to 32.49 nm, the overall median value of the femoral heads combined was 9.55 nm, figure 6. The femoral heads with identified wear can be identified as orange within the bar graph and the unworn femoral heads are shown in blue. From ISO standards, metal surfaces should have an Sa (average areal surface roughness) value <50 nm, however, manufacturers will likely have internal tolerance limits much lower, for example <10 nm [23,24]. It can be seen from figure 6 that the majority of the femoral heads are under or very close to this 10 nm limit. This indicates that the unworn surface zones across the cohort had not been significantly roughened for the majority of the femoral heads.
Femoral head A has the highest median Sa value for its unworn area (32.49 nm), this could indicate that femoral heads with greater wear will have a rougher overall surface, including the areas indicated as unworn from CMM measurement. However femoral head B does not support this theory with a low unworn median Sa value (11.10 nm). A larger cohort is needed, to accurately analyse this phenomenon. Although the median Sa value for a femoral head in the cohort may be low, this does not indicate that the entire unworn surface is 'as manufactured'.
Firstly, it must be considered that the 25 surface measurements only cover a small portion of the total surface, with a high probability of isolated rough areas remaining unmeasured. This is evident within the data collected for femoral heads A and H, as both had a single high anomaly value excluded from their datasets (Sa, 235 nm and 490 nm). Also, surface images indicate that light scratches on the unworn surface are common, however, these do not significantly increase the median Sa value. When comparing the measurements

Overlay plots of worn femoral heads
As the surface measurement locations are precisely tracked in relation to the CMM wear map, the datasets can be combined in a fused output as an overlay plot using the fixture MATLAB software. This allows the surface topography parameter values to be plotted as a colour value overlayed on the wear map. The darkest area on the wear map indicates the wear area, whilst the red circles represent a highest Sa value and dark blue circles represent low Sa values, figure 5.
The Sa values from the grouped measurements are shown in figures 5(a), (b) as an overlay plot. The roughness increases within the wear area on both femoral heads. However, the increase in roughness varies within the wear area depending on its location, particularly for femoral head B.
For femoral head A it was noted that the measurements as shown in figures 5(a), (b) at the boundary of the CMM defined wear area did not necessarily match the surface roughening boundary. This is why stitching stripes of surface measurements provides a more effective analysis of the wear boundary, ensuring a full line of measurements are taken through the wear area. This includes boundary edges and a portion of measurements in the 'unworn' area. The results from the stitch measurements can be seen as an overlay plot, figures 5(c), (d). For both femoral heads, the roughness increases within the wear area compared to the stitch in the unworn area which has no increase in roughness. However, each femoral head has a unique pattern to its roughening. For femoral head B the area of roughening is limited to a small portion of the wear area, whereas on femoral head A the roughening extends out of the wear area towards the pole. This will be shown in more detail through surface topography mapping.

Surface topography mapping
The CMM defined wear area should contain the location of most roughening [7,18], however its shape in the present work does not perfectly represent the shape of the surface topography change. Using an overlay plot of all surface measurements, both the grouped measurements and stitch measurements, the area of roughening can be established and compared the shape of the wear area, figure 5. Clearly a greater number of surface measurements taken allows for a more precise estimation of the differing surface topography areas. Within each defined surface topography area, the median surface parameter value can be calculated, by using only the surface measurements within each area.
For femoral head A, the area of most roughening (Median Sa: 138 nm) is at the lower centre of the wear area, surrounding this is a lesser but significant level of roughening (Median Sa: 94 nm) extending out towards the pole. Underneath the wear area, it is noted that the roughening stops abruptly before the surface becomes smoother again. The area of roughening extending towards the pole would not have been captured without the use of the stitch measurement approach, figure 7.
For femoral head B, the area of most roughening (Median Sa: 118.5 nm) is very small in size and is located towards the centre right of the wear area. The lower level of roughening (Median Sa: 50 nm) is also relatively small in size, leaving much of the wear area with little to no increase in roughness compared to the

Stitched measurements through the wear area
By stitching together surface measurements a continuous topographical map through the wear area can be created. This enables crucial stages of surface development due to wear to be highlighted across the wear zone allowing tribological mechanisms to be investigated. Figure 8 shows sections of stitched surface measurements through the wear area of femoral head A, figure 9 shows the entire stitch line. Figure 10 shows sections of stitched surface measurements through the wear area of femoral head B the central wear zone in this case figure 9(C), shows a different topography possibly indicative of adhesive or oxidative wear and figure 11 shows the entire stitch line.

Metal-on-metal wear mechanism
As manufactured, the bearing surface of total hip replacements are highly polished, average areal surface roughness Sa values are expected to be lower than 10 nm. As demonstrated by surface measurements of low wear surfaces from this study. The contact area between the bearing surfaces occurs on the superolateral aspect of the femoral head due to the natural hip inclination angle [7].
Depending on a wide range of surgical, patient, and implant design factors, the bearing lubrication conditions will be determined. Ideally, the bearing should operate in a mixed to full fluid lubrication regime, as this minimises wear [20,31]. Large MoM implants as examined in the present study were specifically designed to increase the presence of lubrication, compared to the more traditional metal on polymer (MoP) [30,31]. If boundary lubrication conditions do occur, increased wearing of the bearing will initiate through the removal of surface asperities. Other wear forms such as oxidative wear or fretting could potentially be measured with such fixture system with equal success.
For MoM implants the wear will initially be adhesive [31], this is when local welding between asperities occurs, which are subsequently broken in movement creating wear debris, thus leading to the possibility of third body wear also occurring. Wear debris can disperse from the bearing interface and potentially initiate adverse tissue reaction and aseptic loosening.
Roughening occurs within the wear area, with unidirectional scratches typically found at the most roughened location, scratches would appear to be aligned with the natural movement of the femoral head within the acetabular cup, figure 12(a). Outside the wear area, light multi-directional scratches are found on the surface, likely from third body wear debris dispersing from the wear area and interacting across the surface, figure 12(b). These light scratches do not significantly increase the roughness or create further wear debris. These findings are supported both from this study, and previous studies [11,15].

4.2.
Benefits of the custom fixture system 4.2.1. Precise surface topography data capture It has been shown in this study that the area of most roughening is generally within the wear area of the component. However, this study has also shown that the area of roughening does not closely match the shape of the CMM measured wear area, the area of roughening can be smaller than the CMM defined wear area (femoral head B) or extend out of the CMM defined wear area (femoral head A), figure 5. It is noted that the CMM based volumetric wear measurement method proposed by Bills et al [7] on a 60 mm diameter resurfacing femoral head, and considering uncertainty sources, for a coverage factor of 2, resulted in an expanded uncertainty of +/−1.859 mm 3 . In contrast manually positioning the surface maps within a CMM defined wear scar after the part has been physically transferred to an interferometer will only have at best a positional repeatability of several mm and it this not compatible with high precision CMM measurement.
A typical surface measurement method for analysis of the wear area could consist of ten equally spaced measurements within the wear area, using a manual adjustment fixture, this could lead to multiple measurements, or all, missing the area of roughening, leading to an inaccurate analysis as the repeatability of the measurement in terms of the wear scar is of the order of mm. With the developed fixture system, the angular repeatability of the optimised stages was, in elevation, 0.028°and around the azimuth, 0.034°, as measured using a CMM. For the largest head measured in the present work (dia 54 mm) the positional repeatability was consequently found to be 26 um in elevation and 30 um around the azimuth.
Using the excellent repeatability of the fixture system figure 13(a) shows the surface measurements taken on femoral head B. This includes four stitches through the wear area and a set of grouped measurements, this allows the surface topography areas to be well defined. Instead, if a manual fixture is used with poor positional precision, for 10 surface measurements, e.g. Figure 10(b), then the measurements in this scenario may not capture the extent of the roughening, and the area of most increased roughness (red area) could be missed entirely. This leads to an inaccurate surface analysis of the femoral head wear area, with a significant underestimation of the roughening. The average surface parameter values using a manual fixture will as a result not accurately represent the actual development of the topography during tribological wear.

Advanced surface analysis tools
By using the custom fixture system, advanced novel surface analysis tools, surface topography mapping, figure 5, and stitching through the wear area, figures 8-11, was facilitated. To the authors knowledge, there are currently no commercial systems available with this capability. The presently developed analysis tools additionally allow the bearing area surface topography to be quantitatively studied at a level of detail previously unattainable and allows for an improved judgement of hip implant performance.
Surface topography mapping allows the shape of the CMM defined wear area to be compared to the shape of the areas of roughening (or other surface topography change). Crucially this study has proven that the areas of roughening do not necessarily closely match the shape of the CMM defined wear area. Roughening does not clearly define material removal in early stages of wear and may indicate material displacement.
Stitching of a linear band of surface measurements through the wear area ensured crucial stages of topography development through the wear area, such as the border of roughening, could be analysed in depth in terms of surface topography and potential tribological mechanisms.

Faster surface measurement
As the introduced fixture system is motorised and surface measurement locations can be pre-selected on the control software, surface measurements can be taken at a faster rate compared to using manual fixture setups. There is no need for fixture/component adjustment between each measurement, simply the push of a button controls the fixture to move the bearing component to its next pre-determined location. This allows for many more surface measurements to be taken in a shorter time span, when compared to manual fixture setups.

Use of proven metrology instruments
The custom fixture system utilises proven measurement instruments. Instead of creating an entirely new fused metrology instrument, the custom fixture system is utilised to fuse the data from two, tried and tested instruments: the traceable CMM, and a calibrated optical white light surface interferometer.

Quality inspection
Although this fixture was primarily designed for wear analysis of hip implant bearings, it would in addition be useful for quality inspection of newly manufactured components. The software can be used to quickly relocate the implant to pre-determined surface measurement locations that are set by the user. A stitch (or multiple stitches) can also be created across the entire bearing surface to allow for a continuous detailed inspection of the surface topography.

Low-cost design
The custom fixture used off the shelf components and as a result the total cost was below kept low, £3000. This means the design could be replicated or further developed relatively easily in research and industry quality control environments.

System drawbacks and improvements
By increasing the height of the front and end mounting brackets the azimuth rotary stage could be set lower compared to the elevation rotary axis. This would allow larger components to be held by the fixture and correctly lined up with the centre of the elevation rotary axis, this is particularly useful for large heads, up to the currently largest standard MOM resurfacing heads of 68 mm diameter and acetabular cups.
Currently acetabular cups are measurable with the fixture system but depending on their size (68 mm diameter) they could be too large to be aligned correctly with the elevation rotary axis. Assessing acetabular cups using the fixture can establish wear zone limits and offer additional tribological insights to dominant wear mechanisms especially where the head and cup have differing hardness and where wear may only be observable on the acetablaur cup as is the case with metal on polymer (UHMWPE) acetabular cups.
Measurement of acetabular cups requires a slight alteration to the fixture system holding device, and small alterations to the control code would be required. There is also less optical lens access to the acetabular cup surface area compared to the femoral head surface. This would limit the area coverage of the system, however this physical limitation would be the same manual fixturing.

Conclusion
Through the successful development of a computer numerically controlled (CNC) fixture system, a novel surface measurement method for worn hip implant bearings has been created, this can facilitate precise data fusion of CMM and surface measurement datasets. Using surface analysis tools such as surface topography mapping and stitching through the complete wear area allows the worn bearing surface to be quantified and studied at a level of detail previously unattainable.
The presented methodology can allow for better judgment of hip implant performance and can be utilised for both in-vitro (simulator) and ex-vivo (explant) testing for all material types. As the development costs of the fixture were kept minimal, redevelopment across industry or research environments is achievable. Alternatively, instead of use for wear testing, the fixture system can also provide useful for quality inspection of newly manufactured components.

Acknowledgments
The author would like to thank (i) The London Implant Retrieval Centre for providing the femoral components for this study.
(ii) The EPSRC funded Future Metrology Hub for access to the metrology facilities EP/X038513/1.

Data availability statement
The data cannot be made publicly available upon publication because no suitable repository exists for hosting data in this field of study. The data that support the findings of this study are available upon reasonable request from the authors. Figure 13. Comparison of fixture methods a) Custom fixture system. Black area represents the CMM defined wear area and the red areas represent the roughening, defined via surface measurement data. b) Measurements taken using a manual fixture. White circles represent example locations of surface measurements taken within the wear scar. Note only two are taken in the areas of roughening, none capture the main roughening area (red area).