The creation and manufacture of automobiles favor the lightweight, robust, and energy absorbing features of composite materials. Included in this are the braking systems of automobiles.
Brake pads and their lining are complicated materials which are made from four parts including the binder, reinforced fibers, padding, and friction performance modifiers. Analyzing these materials at a microscopic level creates a need to seperate each of the many elemental phases in the material. These analyses are beneficial in finding out the wear characteristics, reverse engineering or failure analysis of such materials.
Elemental mapping in the X-ray microanalysis will demonstrate the elemental distribution in the material, but cannot show the whole story. Complicated materials like brake pads and linings have many obscure material phases which are often individual mineral oxides.
To carry out an accurate analysis of the material, the researcher needs the equipment to characterize each separate phase. Thermo Scientific TM X-ray Microanalysis Systems contain a proprietary principal component analysis algorithm called COMPASS PCA.
COMPASS PCA analyzes the collection of spectra taken from a SEM/EDS acquisition utilizing Spectral Imaging, which provides a spectrum at every pixel in the acquired SEM image. The dataset provides the statistical basis for the PCA analysis.
COMPASS PCA categorizes each statistically individual phase without including elemental composition. This method is explained in detail in a webinar recording named, ‘Unleashing the Power of COMPASS PCA’.
At the moment, there are four known categories of friction brake compositions. These are asbestos, semi-metallic, non-asbestos organic (NAO) and ceramic.¹
As the initial generation of brake pads, Asbestos brake pads have around 40 to 60% of asbestos material within them. The greatest benefit of using asbestos is its low cost. However, due to asbestos materials causing lung cancer, products containing it have mostly been discontinued.
Semi-Metallic Friction Formulations
A semi-metallic brake pad is made from a lining which utilizes steel wool as a reinforcing fiber. Most of these are created from at least 60% steel by weight. The steel fibers perform as the framework to hold the friction ingredients in place.
Non-Asbestos Organic Friction Formulations
Consisting of organic fibers utilized to reinforce the friction materials and give strength to the brake pad, Non-asbestos organic (NAO) brake pads are another option. They were initially created to replace asbestos pads and include less than 20% steel by weight.
Ceramic brake pads have been utilized since the mid 1980s and have no steel fibers. Rather, they use ceramic and copper fibers to control the dissipation of heat.
A huge issue in the creation of automotive brake pads or shoes is managing the ingredients and distribution of such components. Reproducibility of the formulation and distribution of raw materials is important in creating sound quality control in the final pad or shoe.
X-ray microanalysis in scanning electron microscopy (SEM) utilizes energy dispersive spectroscopy (EDS) to define the composition of materials in the brake pad. Adding a statistical method to this analysis can be used to understand the distribution and phases of each component in the brake pad or shoe.
Materials and Methods
The brake pad used in this analysis was worn and therefore had been through various heating and cooling cycles, including exposure to harsh weather conditions and salt used to melt snow on the road.
A sample of the pad was taken from the backing plate and embedded in epoxy, measuring approximately 1 inch by 1 inch. The embedded sample was then polished using various grit sizes with a last polish of 1 micron grit. The sample which was not coated was analyzed on a JEOL JSM-7001F scanning electron microscope (SEM).
The EDS system utilized to collect and measure X-ray data was a Thermo Scientific™ Pathfinder™ System with a 10 square millimeter silicon drift detector. Spectral Imaging data was taken at 20 Kv accelerating potential, 30x magnification and at a pixel resolution of 1024.
The SEM examination of the sample illustrates many phases of the brake pad material with some fairly large phases (shown in Figure 1).
Figure 1. SEM image of the brake pad sample.
An x-ray analysis of the sample demonstrates that there are 16 elements present (see Figure 2). A few of the elements such as Fe, O, and C, are most frequently present. The existence of the Zn K line peak shows that Zn is present. However, the Na K line crosses over the Zn L line which makes identifying the element Na complex.
Figure 3 illustrates the elemental X-ray maps. To look into the spectral imaging data further, a red, green, and blue (RGB) overlay of elements was utilized to determine some of the compounds that may be present. Because Ba and S was present in the sample, the initial compound to explore was BaSO4.
The yellow particles (which can be identified by red and green which create yellow) show the presence of Ba and S in the same locations. Therefore, Ba and S and most likely in the form of BaSO4 (shown in Figure 4). The red particles shown in Figure 4 show that S is present in some other compound. Most of the elements are present in very small concentrations and are sparsely distributed.
Figure 5 presents the distributions of Na, and Zn (yellow particles), which allows one to conclude that Na might not be present. The acquired spectrum of the bigger yellow (Na and Zn) particle (as shown in Figure 6) demonstrates that the Zn L peaks are present but the Na K peak is not, and suggests that Na is not present where Zn is present in the sample.
As there are many elements of light in the sample along with a varied distribution of O, a large number of these elements could be oxides which can be complicated in terms of composition. Due to the large number of elements, the utilization of the RGB technique could prove to be lengthy and drawn out.
COMPASS principal component analysis (PCA) was used instead to favor the mathematical approach and to analyze the sample in more detail.
Figure 2. X-ray spectrum from pad sample.
Figure 3. X-ray maps showing the element distribution from the brake pad sample.
Figure 4. Distribution of Ba and S.
Figure 5. Overlay of Zn and Na X-ray maps onto the gray reference image.
Figure 6. Extracted spectrum of Na/Zn particle region.
COMPASS PCA Results
Area method and Spectrum method are the two available processes when using COMPASS PCA. The area method places importance on the area of each component, where the spectrum method uses the spectrum to locate the components.
The results obtained from the COMPASS PCA shows that there are 21 components when using the area method (see Figure 7) or the spectrum method. Upon looking at the component images and their corresponding spectra, some components can be mixed together such as C2 and C20 being the carbon component (shown in Figure 9).
After viewing the spectra from each different component, some of them are equal in composition and are different only in terms of the concentration of the elements. Comparing the various components shows that only 13 distinct components exist. Entering the total number of components as 13 in COMPASS PCA, and re-extracting the components provides 13 distinct components (illustrated in Figure 10).
Using the component maps, the researcher can see that C6 is the Zn component (shown in Figure 11) and C12 is a FeCuS component (Figure 12). Each are preserved as distinct components during analysis. After this, the job is to look at each component spectra and find whether more can be combined.
C1, C7, and C11 all demonstrate similar elements (see Figure 13). Although, component 1 has a larger AI content than C7 and C11. Also, C1 and C11 both have Zn present where C7 lacks this element. Comparing each component in this manner takes quite a lot of time.
A simpler and more efficient way of comparing components is by using the COMPASS engine and its utilization of statistics to mathematically discover the various phases.
Figure 14 demonstrates the phases extracted from the reduced 13 COMPASS components. Analysis of the spectra and distribution of each phase illustrates the composition of the brake pad (Figure 15).
Overlaying the ten phase distribution images onto the reference electron image demonstrates the separate distributions of each of the phases (Figure 16).
Figure 7. Principal components of the pad sample using the area method of analysis.
Figure 8. Principal components of the pad sample using the spectrum method of analysis.
Figure 9. Comparison of Component 20 and Component 2 showing similarities in the C components.
Figure 10. Extracting the X-ray maps limited to 13 unique components.
Figure 11. Component 6, Zn rich component.
Figure 12. Component 12, FeCuS rich component.
Figure 13. Comparison of spectra from C1, C7 and C11.
Figure 14. Phases calculated from 13 components.
Figure 15. Phases analysis showing the composition of the brake pad.
Figure 16. Overlay of the 10 phases onto the gray level reference image showing phase distributions.
The automotive disc brake used in this analysis is a semi-metallic type of pad which is shown by the large iron content discovered in EDS analysis. Along with Fe, there are 15 alternative elements that contribute to the brake pad formulation.
This specific sample is highly complicated in composition which makes it harder to analyze. Utilizing the RGB method of overlaying the element maps would take a lot of time. However, by favoring the statistical approach, the analysis was simpler and took less time.
The method of using Spectral Imaging to obtain element maps, COMPASS PCA to determine important components and then using the components to discover phases was more efficient than using X-ray maps alone.
Moreover, by using this approach there was much less human bias introduced during analysis. This article illustrates a quick and thorough approach when analyzing complicated materials that contain a wide number of elements forming several compounds, through the use of Spectral Imaging and Principal Component Analysis.
References and Further Reading
This information has been sourced, reviewed and adapted from materials provided by Thermo Fisher Scientific – Materials & Structural Analysis.
For more information on this source, please visit Thermo Fisher Scientific – Materials & Structural Analysis.