The Hidden Variables in Your Test Results

[00:00:01] Speaker: We're going to get started. Just to introduce myself, I'm Casey Willis. I'm the strain product manager, like Nick said, here at Instron. Today, we're going to primarily talk about non-contacting strain measurement. But if you do have any questions about other types of extensometers, uh feel free to put those in the chat and if we can't answer your questions today, we'll make sure that we follow up um via email after the webinar. So to quickly go over what we're going to talk about today, I'm going to start with a small introduction, then we'll dive into those hidden variables in your test results um that you might not know about. So those three key points are your hardware configuration, your working distance, and environmental conditions. Uh we'll wrap up with environmental conditions, and then we'll jump to a Q&A. So, we started off the webinar with a quick poll. Um, and it seems like people typically fall into these three categories of extensometers. So, your clip-on, your automatic, uh, non-contacting, and your automatic contacting extensometer. Um, clip-ons are definitely the most common and historically the most traditional type of extensometer, and they can be found in nearly any lab that I visit. Uh, when you're purchasing a clip-on, you're usually targeting a specific gauge length or elongation. If you need a wide range of gauge lengths or you're trying to test to a lot of different applications, you'll likely find yourself purchasing multiple devices. As a contacting solution, the operator has to manually attach and detach the clip-on from the specimen. Um, which can be kind of cumbersome depending on how many tests you're doing a day. Some factors to be wary of when you're actually removing your extensometer during a test is what type of environment are you retrieving it from. Maybe you're testing in a temperature chamber at an elevated temperature, or you could be taking that clip-on off of a loaded composite. So things to just keep in mind. Um, if you do happen to forget to remove a clip-on before an explosive break of a specimen, you could find yourself um facing a high price tag of some maintenance costs or you could be coming back to the manufacturer to ask for more knife edges. I know that's pretty common. Um, people come back and purchase more knife edges if they wear or if they do accidentally leave the clip on uh during failure. Um, so if you don't want to deal with the pains of clip-ons um and the high potential maintenance cost and you are looking at a different um kind of strain solution for a wider variety of materials or gauge lengths, you could be looking at more automatic strain solutions. Uh so we have a contacting solution like the AutoX on the right hand of the slide um and a non-contacting solution like the AVE in the middle. Um and choosing between these two, you should really have that decision be driven by the application that you're testing to. So we talked about those variety of gauge lengths and elongations. Um, but where the AVE might have an advantage over the AutoX are in situations um where you might be testing delicate specimens, those thin specimens could break prematurely from the contacts from the knife edges. Um, so if you see any of that or you're primarily testing thin films or foils, going the non-contacting route um is a pretty kind of key decision maker. Um, on the other side of the material spectrum, let's say you want to test a very strong material with an explosive break. Um, you could hypothetically use an automatic contacting solution for this, but you run the risk of increased wear if you're actually holding that specimen on through break. So, you're probably want to keep want you probably want to keep spare knife edges or part replacements um on the shelf just in case you have to maintenance your device. Um and then on the product versatility side of things, either a non-contacting or automatic contacting device could be configured with transverse options. Um the AVE has the ability to average 32 strain um transverse strain measurements which could be really impactful if your lab saw variation in R value. Uh the AV is also preferred for chamber applications as it can look through the glass on a temperature chamber um and track strain through failure. The AVE can also be configured with other advanced options like uh 2DIC. So either way, if you're looking at an automatic um strain solution, you'll certainly see a throughput increase for your lab compared to those traditional clip-ons.

[00:05:11] Speaker: So regardless of what type of extensometer you're testing with, the standard you're testing to is likely pointing to ASM E83 or ISO 9513 for compliance. So ASME E83 and ISO 9513 are the standards for guiding the practice of verifying and classifying extensometer. So E83 is published by ASM International which is most common in the US while ISO is a little bit more common worldwide. Um, but high level, even though they're a little different, the high level goal of just placing a standardized amount of guidance around dictating the classification of your device is very similar. So, we'll dive into E83 today. Um and that focuses on ensuring that extensometer including non-contacting types like the AVE meet specific performance requirements in terms of accuracy, linearity, and repeatability. So, something that you might not be aware of if you've used traditional clip-on extensometer is how different types of transducers or varying types of technology are susceptible to different types um of noise sources. So, the three categories that are typically a source of noise in strain measurement are those three listed at the bottom of the slide. So, mechanical, electrical, and environmental. At the end of this presentation, we're going to really hone in on environmental because that's going to be really important for non-contacting. But when you're thinking about those three categories, you can ask yourself questions like, what kind of noise is the cable that connects my extensometer to my test system introducing? Is your extensometer isolated from your lab environment? Is it isolated from vibration? And for optical extensometers especially, what can that video extensometer see that you might not necessarily be able to see?

[00:07:21] Speaker: So ASM released a new version of E83 which included annex A1 which is specific to the verification requirements of optical extensometer. And they did this because they acknowledge that optical extensometers behave differently than traditional contacting extensometers and have unique considerations that need to be kept in mind when you're verifying and using those devices. Um so the annex advocates and emphasizes for the importance of setup consistency. So if a service engineer is on site verifying your device, how are we, Instron, the manufacturer of the device, ensuring that your entire setup is going to behave the same exact way between your annual verifications? There's three really important factors here um that influence device performance. The first one, hardware. How repeatable and reproducible is the hardware of your device? What changes could you potentially be making to your hardware and how did the manufacturer, us in this case, account for those effects? Uh secondly, working distance. This is the distance between the specimen surface and the device. What happens if this changes? Let's say a change in specimen thickness. What does that mean? Um, and lastly, environmental consistency, which is likely something you're not aware of because there is a whole lot happening in that pink space between your device and your specimen uh that you're likely um you likely were never told about. So, we're going to start off with hardware considerations. Um, hardware considerations. It's really the first and most important decision you'll make when you're purchasing a device. What type of lens are you going to use during your test? Uh, so you can configure the AV with a 6, 9, 16, or 35 mm lens depending on your testing requirements. The 6 millimeter lens is our longest field of view and you can see it on the bottom left of the slide. Um, and the top kind of looks rounded like those traditional fisheye lenses. Um, if you've ever taken a picture with a fisheye lens, you know it captures a panoramic view of an image. Um, a trade-off to being able to capture such a large amount of field of view in a singular image. That trade-off is distortion. The results will look distorted. Um, so to show you an example of this, the image in the middle of the slide um is a picture taken of this linear checkered board pattern um with a fisheye lens just to help you visualize how the distortion actually looks. Um that linear pattern gets bowed and if you notice at the edge of the field of view of the lens, the distortion actually gets worse. So, you might not know um because it's not really apparent to the naked eye. If you're taking a picture, even with a lens on your phone, um there's going to be some level of distortion, even if you can't visually pick up on it. Um it's semi- apparent with the 6 millimeter lens, but not apparent at all with the 35, a really short field of view. Um, but we do acknowledge that every lens has some unique level of distortion and we needed to compensate for that to really unlock the next level of accuracy with the AVE3. Um, so we introduced a process that essentially teaches the AVE um what the distortion of your lens looks like at different points of the field of view. Um, so this is a file that gets saved and stored on the AVE and when you're switching between lenses, the AVE will be able to reference the appropriate lens distortion file as you're testing to make sure that your results are distortion-free.

[00:11:32] Speaker: Another important hardware design of the AVE3 that really supports that theme of repeatability and reproducibility was the switch from the traditional screw on lens connection to the new kinematic mount. So, the new kinematic mount is magnetic and it does a really good job of ensuring that the lens is installed and removed. Um, when the lens is uninstalled and re or when the lens is installed, it's going to be put back in the same exact orientation no matter who's installing it. Um and this is really important because we didn't want um to jeopardize the focal length of the device. So the focal length is the distance between the image processor and the lens. And we wanted to make sure that that was fixed and consistent during testing even when people were swapping out lenses. Um because focal length is very crucial um in optics. It dictates focus, diverges light, influences your image magnification, and affects your field of view. So, this is not something that should be tampered with. Uh along with a fixed focal length, we're preconfiguring focus and aperture settings um at the factory in Norwood with a specialized tuning algorithm to make sure that those images that the AVE3 is pulling in um are very consistent and it's very standardized.

[00:13:07] Speaker: And then just a quick slide on working distance. um since there's not a whole lot to say but still an important part of ASME E83 in the guidance that that gives um working distance can be defined as the physical distance from the surface of the specimen to the optical extensometer. While it's not important necessarily where the physical distance is measured from, it's important to measure that distance and report exactly where you took those measurements. Um doing this will enable the user to really recreate those conditions under which the optical extensometer system was calibrated to. So for Instron specifically we look at the mounting distance. So where the AV is sitting on the rail um in relation to the face of the specimen. So unless you're testing only version very specific thickness of specimen, your working distance is likely to be a range rather than a single value. So if your specimen thickness changes, we ensure performance is unchanged. Um, and we do this with Blue Hill Universal. So Bluehill will actually use the specimen thickness that you input during a test to optimize and compensate for changes in your working distance. Um, and this feature is used during your calibration procedure just like ASME83 outlines. So we did extensive testing on this and as a manufacturer we should be telling um our customers and the users of our device the range of working distances that this device can be used at because consequences for exceeding the specified working distance range means that you could jeopardize what classification your extensometer could meet. So, you should always be checking with your manufacturer to see what guidance they give around what working distance or working distance range you can be using your device at.

[00:15:12] Speaker: And moving on to environmental consistency. This is definitely the most important factor that influences device performance. And at the end of this section, we're going to look at what that means for your data. We'll have a data uh comparison at the end. Um I just want everybody to take a second and think about what your lab looks like. So on the left, you can see that there's very I pointed out a couple of things in the image. There's very harsh artificial light directly over that test system. There's an HVAC system directly over that test system. And the whole wall of the lab is windows. So when you have windows in your lab, your lighting conditions and illumination conditions in the lab are not constant. Um in the morning you can see very little light. In the afternoon the lab could get really bright. Um that HVAC system, depending on what uh season it is, you might notice that it's blowing either hot or cold air onto your test system. So, what are those variables actually doing? And can your extensometer pick up on those? Um, and this is extremely common. I think all of these things I could take any picture of anybody's lab and call out some of these just like normalized um variables that we see in labs. It's the same thing as the apps lab in Norwood. We have a whole wall of windows. Um but it's important to realize that you need to do um some things on the optical extensometer side to actually compensate for um the variation that these actually introduce to your tests. So focusing on lighting first. So think of that back wall of windows, think of the artificial lighting. Um because lighting really dictates what your camera sees. So every pixel on the AVE3 and its intensity matters since we're interested in subpixel accuracy. So illumination is really the backbone for non-contacting measurement systems and fine control of this illumination is really needed for metrology grade results. Um if illumination is not set correctly, systematic errors may occur. So all image processing algorithms are susceptible to this. Um so it's really important to keep in mind um it's critical to control what type of image your device is inputting. So if you look at um the visual we have on the screen um this contrast gradient there's many uh operator related variables that we introduce um to our system. So what color is your specimen? What color uh marker are you using to mark it? um what is the angle of your light bar? All of these things affect how much contrast the AV will pick up on. So that clear specimen on the bottom right um that is a very bright and crisp dot with a lot of contrast between the specimen backing and the mark. Um compared to on the bottom left, this mark was generated with a very old marker that was on a shelf for probably a year or two. And you can just see this mark looks very muted. Um, so in order to try to just even the playing field and get that sweet spot of contrast, the AVE3 light bar will dynamically adjust illumination conditions until the optimal amount of contrast is generated and then your test will start. So that just removes any of those uncontrollable factors like how new your pen is, what color specimen you're using, um, and you're testing with.

[00:19:11] Speaker: So a aside from what you can visibly see and that you can see on the UI of your video extensometer, um, the truly hidden variables in your optical testing are micro mirages. So, a mirage is an optical illusion caused by atmospheric conditions. On this slide, figure one is a reflection that kind of looks like puddling on a road on a hot day. Um, and figure two is haze generated um when a jet takes off. And these are two examples that you might be most familiar with. Um, so figure one, everybody's probably driven on a hot day. Um, you might notice that those puddling reflections as you're looking forward out on the highway or on the pavement. Um, and you know there's no water on the road, but why can't you see the pavement? So, the refraction of light through layers of air with different temperatures and densities is actually causing this effect. So, light rays from the sky are bending or refracting as they pass through the cooler and denser air from above into this hotter and much less dense layer causing the brain to interpret this as a reflection of the sky on the road. Um, and it creates that puddle illusion. So, this interaction of light rays and varying air densities and temperatures is present on the micro level which is really important to acknowledge. and submicron devices like the AVE3 can see these. Um, figure three is a visual from NIST that they created during COVID to help everybody visualize where the air from your lungs goes when you breathe, talk, or cough. Uh so you'll notice in this visual a very similar optical illusion occurs as um you can see in figure 2 with the jet engine that kind of haze that's difficult to see through. Um keep that in the back of your mind um in the next two slides that kind of hazy effect. Um, so if you take a look at figure 2, if you tried to take a measurement of the height of the building in the background of that image at different points in time, um, you're likely going to get different results. Um, because of that mirage ebbing and flowing, some of your displacement measurements might be slightly up or slightly down. Um, and that principle is essentially what happens on the micro level with video extensometers when they're when they're being subjected to these temperature and density changes in the lab.

[00:21:53] Speaker: So, we know laboratory environments are very dynamic. Nobody is sitting very still in front of their system testing all day. Uh there's HVAC systems, there's people walking, there's people talking. Um so a quick way to visualize what's happening in that airspace between your extensometer and your specimen is to perform a smoke pen test. So you can light the end of your smoke pen and allow that smoke to just enter the test space. Um you might notice uh that the some portions of the field of view the smoke is puddling while others it's kind of flowing away and that can help you visualize the air densities and how those are interacting. Um to see what this looks like to the video extensometer we performed an analysis which can be seen on the left bottom portion of the slide. So this analysis is widely used for fluid dynamics and heat transfer research because it makes invisible phenomenon’s visible. And in this case, our engineering team was trying to capture data to showcase just how much mixing of different air densities um is occurring in front of your video extensometer if you do nothing to control for environmental um disturbances. Um, and these disturbances can manifest as either large or small signal changes to the extensometer. And you can see in that um kind of heat map as everything's moving through um where we place those targets which are marks on the specimen um the resulting analysis shows a very inconsistent and varying map of strain uh signal changes over your points of interest which are those dots. Um, and these signal changes can be amplified if we introduce something like an HVAC system turning on. So that kind of heat map can look a lot worse.

[00:23:55] Speaker: So this is something we really wanted to kind of hone in on the AVE3. Having those large and irregular signal changes becomes a big problem if you want to run a test in strain control because you'll probably have to use excessive filtering to hide those large strain signal spikes. Um, and your system will likely lag behind and not be able to respond to those real-time strain events. or you're going to be forced to use a very large averaging interval and you could run the risk of not being in compliance with a standard um or seeing unwanted variation in your yield properties. So to avoid all this uh we really attacked the core of this problem because we knew we wanted to reduce what was contributing to our extensometer just baseline level noise profile. So with the new patent pending airflow design and our constant density air technology, the AVE counteracts this Mirage effect to give us a very stable um and reliable strain signal. So if you look at that same smoke pin analysis on the top right of the slide, um you can see that there is no pooling of the smoke that you saw on the previous slide. um and the entire field of view is just very turbulent. Nothing can hang around um and interfere with what the AV is seeing. So that actually results in a much more consistent image being fed into the AV. Um if you look at the Schlieren analysis on this page on the bottom left, you see a lot more blue. The majority is blue. It's very constant. There is just a constant small signal change. And something that is very consistent is much easier for us to compensate for than signal changes that are very large. Um so it's just really important that you're protecting your test space and um protecting the full field of view of your extensometer. So, this is a visual from our overview video to just help you see how the AV is just capturing air and pushing that through um the test space to make sure let's say those arrows are a heater kicking on that any of those negative effects are just bouncing off um of that condition test space.

[00:26:35] Speaker: And then to wrap this up, what does this look like in terms of data? Because that's ultimately what matters. Um, during some comparative testing with a standard video extensometer with no environmental mitigations and the AVE3, which has that new airflow pattern that we just talked about, um, an engineer noticed that when he was standing next to the standard video, every time he spoke, he would see a jump in the strain signal. Um so obviously somebody talking is not a very standardized test. So we wanted to get a very apples to apples um study performed on this. So what we did was we set up a hot plate next to these two extensometer. An extensometer that has no sedate equivalent no environmental conditioning um versus the AV. So, we introduced a one degree C air disturbance, which is essentially like a heater kicking on um into the test space. So, we gently blew um air from a hot plate into there. And you can see on the left there was a huge noise kind of jump for that standard video extensometer. So, a 40% increase in noise. You can see the spikes of 6, 7, 8 microns level of error in that measurement. um while the AVE had no loss of accuracy. Um so being immune to these just totally normal disturbances that your lab can be subjected to um is really important if you want that micron level of accuracy. Um if you don't do anything to shield your test space from this, you could be out of compliance with a standard that dictates a very specific amount of allowable error. uh without actually knowing. Um so it's good to always protect your data integrity um protect the field of view of your video extensometer. Um,

[00:28:41] Speaker: and then if you're interested to learn more about the AV unit and its features, or you're interested to see in real time in a product demonstration what the effects of that new airflow pattern have um, when the fans are on or off on the AVE. Feel free to scan these QR codes. So you can take it you it will take you to the product page and the website um or you can just watch the overview video that we have um online and then Nick I think I'm handing it back to you for this slide. All right. Well before we get to the upcoming webinars um we do have some questions to throw your way. All right let me stop sharing. Um so we'll go through this first one. Um has the has accuracy changed from the AVE2? Yep. So accuracy has improved from the AVE2. Um you can see it specifically in the repeatability of modulus measurements. Um but ultimately if you take a look at the POD you'll see lower gauge lengths getting that kind of 0.5 um or B1 classification. Um so better performance at lower gauge lengths and then just more reliable performance throughout whatever gauge length or elongation that you're testing with.

[00:30:15] Speaker: Um, is the AVE3 able to test in strain rate control to um, ISO6892? Yep. So whether it's ISO6892 or ASME8, um, unlike the AB2, it's recommended to test in strain rate control with the AVE3. Um, kind of what enabled this I would say with the AVE3 were the improvements that we made to the airflow design and the CDATS and just really minimizing the amount of noise on our strain signal um because we wanted to be able to have the extensometer um tell the frame to react to very real-time strain events. So to do that with very little filtering, we needed to really lower our baseline level of noise um of the device. All right. Um is there a way to turn the fans off on the AVE3? Yep. So that was definitely something customers um especially thin film customers uh struggled with the AVE2. Um specimen loading. You can imagine if you try to hold up a really delicate specimen in front of a fan, it would kind of just blow out of the grips. Um, so with the AVE3, we do have either a toggle button where you can just override the fans to be off all the time. Um, or preconfigured settings to only have the fans turn on while you're testing. Um, so it's really up to the user on how they want to configure um, their fans and what settings they want to use. All right. Um, can we still buy the AVE2? So, the AVE2 for the time being is only available for dynamic systems. Um, so for any static applications um, for those universal test systems, you would just be quoted in AVE3. All right. Uh Casey, I'm going to give you a second to cruise through um some other questions and I'll jump into some of the recap and things that I want to make sure um I cover. So um for just a reminder people that weren't here when I first started. So we did we are recording this event um and I will be sending these out uh a link to this. It'll be on your YouTube. Um so anybody that registered will receive a link to the recording. Um, and then also Casey, if you could do me a favor and put that webinar slide up for the upcoming webinars. Yep. I think that's, you know, given this topic, I think there's probably a couple other topics coming up in the next month that some of you might be interested in. Um, actually going to drop a link into the chat real quick for people to easily access these like if you if you want to sign up. Um, all right. The slide should be up. Okay, cool. So, there's a couple links there. One you can click on it looks like. And then, so that the first one is for our webinars page. It has all of these sessions. So, if you have any interest in these topics, sign up there. And I also would suggest you sign up um subscribe to our YouTube channel just because, you know, people like Casey and our other apps engineers and product managers are putting out great content all the time. Um, that's really the best place to get it. We try and put as much content out on our YouTube channel so you have access to it whenever it's uh published. Um so now I've gone through that Casey, I don't know if there's any other questions um you want to get into here. Otherwise, I know we can just follow up with individuals on more use case situations. Yeah. Um no, we'll definitely follow up because I know there was one uh question about okay, what's the biggest improvement from AVE2 to AVE3? Um, and I feel like to accurately talk about that, I would want to learn more about your application and what you're doing in your lab. Um, a really quick question we can answer. Can uh I think somebody asked about the lenses and if you can put the new kinematic mount on, um, the AVE2 with the screw style lenses. Let me know if I've read that question incorrectly. Um, but the answer is no. So, you would need to purchase an AVE3 to get that new um housing and new kinematic mount design to be compatible with those new um magnetic lenses. If you watch the overview video, I think you can visualize just how different the housing body in the Ave3 is from the AVE2 screw on style connection. Um, just so you can learn a little bit more about what that kinematic mount looks like. Um, that being said, the old style lenses are still available. So, if you are an AVE2 user, you can still purchase those. Um, and it's not like it's not like those are going away.

[00:35:15] Speaker: All right. So, I think this is a our stopping point. Um, before everyone goes, um, I do want to mention there's going to be a short survey that appears once our webinar ends. We would really appreciate I know I would really appreciate if you take a moment to share your feedback. It helps us improve future sessions. Um so thank you Casey for great presentation and certainly thanks everybody for attending. Um take care and we hope to see you again uh see you again soon.