Crash testing initiative on actual road surfaces

Discussion in 'Equipment' started by ukAdventurer, Mar 3, 2021.

  1. ukAdventurer

    ukAdventurer Long timer

    Joined:
    Jul 9, 2007
    Oddometer:
    1,108
    Location:
    Ulaanbaatar, Mongolia
    I recently took early retirement and I've decided to do some crash testing of rider apparel on actual roads. This is a non-commercial, non-profit initiative for motorcycle (and similar road sports) riders' benefit. I've already started and I'm in the process of collecting samples of commonly used materials and some garments all bought from my own funds. I've designed and built a simple mobile drop tower. I've also got simple concepts of mobile open drop testers for irregular shaped garments (gloves, boots, larger samples of jackets/pants). By taking the tests to the environment I could test on roads and some off-road surfaces possibly.

    Here are some very short videos of design, build and first test.

    So far I've got these materials:
    slip/comfort liners (used inside jackets)
    polyester mesh (heavyweight for ventilation in outer panels)
    nylon 500d (woven coated lightly)
    polyester "Oxford" 600d (woven coated)
    nylon cordura 500d (coated)
    nylon cordura 100d (woven uncoated)
    denim (14oz)
    aramids (jersey knit 180 gsm, 200 gsm, 340 gsm)
    Covec/Vectran (jersey knit SR6 - 6 refers to seconds slide time)
    cowhides (1.4mm to 2mm)

    Bought but not received yet:
    Schoeller dynatec
    Schoeller dynatec with ceramic plates
    Schoeller Keprotec
    Competition grade cowhide

    Trying to source:
    denim blends used in the single layer jeans
    1680d (or higher) nylons and polyesters

    One of my motivations is to see if certified gear (materials and whole garments) is fit for purpose. I'll test and rate how gear will actually perform in the real world. This is especially applicable to textiles. The root cause of failures being seen now seems to be EN 17092 and EN 13595 don't include a realistic drop height during their abrasion tests. The drop heights are currently 10mm (0.4 inch) and 50mm (2 inches) respectively. Essentially no impact loading or very little, compared to reality. Who falls from 10mm to 50mm?

    Below is a link to a project to improve testing and fix EN 17092's AART (Darmstadt test) which can not do drop tests from realistic heights.

    http://pioneers-project.eu/wp-content/uploads/2021/02/Deliverable-D3.1.pdf

    So the problem is widely understood and it's being worked on but bureaucracy takes a long time. So, hopefully I can assist riders in making informed decisions and asking the right questions when choosing gear.

    If this is interesting to you please leave a comment, suggestion. I'll attach some images to show my work so far. I still need to tweak the drop tower a bit but it's pretty close now.

    One question I'd like to ask members: what would be a suitable platform for the results of this research and testing? Youtube? A dedicated website? An existing website? Or....?

    test of denim (no layers, no armour)
    25kph - 5cm drop height - 1 layer of denim.jpg
    test of denim (no layers, no armour)
    35kph - 50cm drop height - 1 layer of denim.jpeg
    a single Frame from the GoPro video of the first drop test (improvements needed to avoid rotation of the sample)

    Drop test - 1 frame.png
    drop test of 2 layers of cowhide (non-specific to motorcycling)
    drop test cowhide.jpg
    tempImageU9OEYA.jpg
    the road surface for leather test (very rough asphalt)
    IMG_1936.jpg
    #1
  2. Eric80

    Eric80 Been here awhile

    Joined:
    Sep 26, 2020
    Oddometer:
    147
    Location:
    Maine
    Can you explain (or shows pictures/videos of) the testing procedure? If we are thinking about real world performance, I'd be interested to know more about the pressure, speed, and distance of the abrasion test.
    #2
  3. Kinsman

    Kinsman Ribs....for her pleasure

    Joined:
    Feb 11, 2005
    Oddometer:
    6,176
    Location:
    Stevensville, Montana
    You'd need a 190 lb dummy.
    #3
  4. Suburban_Hobo

    Suburban_Hobo Can you point me to the nearest Starbucks?

    Joined:
    Aug 14, 2019
    Oddometer:
    92
    Location:
    Atlanta, GA
    There should be plenty of candidates in CSM.

    :hide
    #4
    lvscrvs, JeepDawg, Tantor67 and 11 others like this.
  5. SmittyBlackstone

    SmittyBlackstone Long timer

    Joined:
    Mar 3, 2015
    Oddometer:
    1,657
    Location:
    Westchster County, NY
    Did somebody call me?
    #5
    SidewinderX, Josh69 and ukAdventurer like this.
  6. Kinsman

    Kinsman Ribs....for her pleasure

    Joined:
    Feb 11, 2005
    Oddometer:
    6,176
    Location:
    Stevensville, Montana
    :drums
    #6
  7. ukAdventurer

    ukAdventurer Long timer

    Joined:
    Jul 9, 2007
    Oddometer:
    1,108
    Location:
    Ulaanbaatar, Mongolia
    I'm still deciding on a protocol - the rig is flexible enough to choose any protocol. Lots of folks are already testing abrasion times/distances but they are missing realistic impact.

    For that reason, I'm leaning towards just a first impact test one zone 1 (joints) with a 5kg drop weight (with a curved surface striker - similar to humans) from 1 meter producing a 50 joule impact. This is similar to what EN 13595 used for impact protectors use. 1 meter and 5kg is a middle ground for knees, hips, elbows and shoulders which all have somewhat different masses, loads, pressures. Static loads don't really apply because it's mainly an impact test where the striker will bounce and settle several times (usually 3 times).

    I will use an armour pad with all tests, if the garment assembly includes one. So, I'd estimate in the case of an elbow simulation, a static load of 25 Kpa over a about 2000 mm2 area. (50mm diameter static area with 5kg).
    I may do a drop and constant speed until failure (EN13595 style) or a drop and stop (EN 17092). I'm not restricted to any one protocol but I'd like to decide on one that is most useful.

    My own thought is riders would like to see if their gear could survive multiple low-speed accidents compared to the "one and done" school of thought with textiles. What do you think is most useful test for riders?

    Here's a link to a playlist of concept (3d animation), build, and first test. The video of a first test shows a 50mm (2" drop) at 15mph. The single layer denim burst on impact (photo in original post).

    I'm improving the drop test so the weight doesn't rotate on any axis. It should be able to rise up and down though.
    #7
  8. ukAdventurer

    ukAdventurer Long timer

    Joined:
    Jul 9, 2007
    Oddometer:
    1,108
    Location:
    Ulaanbaatar, Mongolia
    I'm 185lb (84kg) but not willing to throw myself off a bike repeatedly :rofl If I was a rich man, I'd get a Hybrid III dummy but you can do a lot of testing on first impacts without any dummy. Future possibilities are load sensors, thermocouples and accelerometers. Maybe some pigskin (raw) and bones.
    #8
    9w6vx and SmittyBlackstone like this.
  9. ukAdventurer

    ukAdventurer Long timer

    Joined:
    Jul 9, 2007
    Oddometer:
    1,108
    Location:
    Ulaanbaatar, Mongolia
    I'm not sure if your comment was serious but you got me thinking.

    Would this work?

    Rescue randy mannequin.png

    It's not a Hybrid III crash test dummy but for impact abrasion it could work. Seeing a crash with a dummy seems to be more visceral than just a sample on a 5kg weight. More importantly nobody gets hurt, except Randy.

    Even better, he's about my height and weight, so my gear should fit.
    #9
    SilentRay, Kinsman and Snake Oiler like this.
  10. ukAdventurer

    ukAdventurer Long timer

    Joined:
    Jul 9, 2007
    Oddometer:
    1,108
    Location:
    Ulaanbaatar, Mongolia
    One question I'd like to ask members:

    what would be a suitable platform for the results of this research and testing? Youtube? A dedicated website? An existing website? Or....?
    #10
  11. Snake Oiler

    Snake Oiler If the world didn't suck, we would all fall off

    Joined:
    Apr 22, 2016
    Oddometer:
    5,572
    Location:
    Pennsylvania, but living in a toy hauler full-time
    I'd think YouTube would get the most attention, but not sure it's the best way to get the information out to people.
    #11
    ukAdventurer likes this.
  12. Jan from Finland

    Jan from Finland Been here awhile

    Joined:
    May 2, 2005
    Oddometer:
    976
    I think that the reason for low drop height might be rebound. Actual sliding time in an accident is only few seconds. About 4 to 7 seconds from a high speed MotoGP crash (at 250 - 300 kmp/h). If you drop the striker from 1 meter, it will rebound and could be airborne more than half of the test time. The test would not be repeatable nor valid, if you can't control how the striker bounces.

    Another issue with your test protocol is that it's quite often a seam which bursts. Not just the material.

    If you do any tests, I will surely watch them on Youtube.
    #12
    Xchemie, Gondul, levain and 1 other person like this.
  13. ukAdventurer

    ukAdventurer Long timer

    Joined:
    Jul 9, 2007
    Oddometer:
    1,108
    Location:
    Ulaanbaatar, Mongolia
    Yes, I think so too. My thoughts are: publish videos quickly, create an article (mcgearhub.com), and create a paper if the findings are relevant to the industry.

    Thanks for responding.

    I agree. Labs don't like high-drop heights because it makes the slide time variable IF you are doing a pure abrasion test. However, that doesn't mean we should ignore high drop impacts. I feel slide only (basically what EN17092 does) is not valid. You won't have a slide with no impact as an every day rider in city, country, highway. MotoGP riders might have the least drop in a low side but that's not how most riders fall off in non-professional crashes.

    One way to make the test repeatable and valid, is to separate the drop from the slide. So a test might just involve the first drop from 50-150cm, allow the sample to bounce and catch it in the air. This should be absolutely repeatable and controllable. Obviously, 4 - 6 tests with similar results would prove repeatability.

    Furthermore, since my tests aren't tied to an interest in a particular test rig (CAM/DAM-AART), I can test whole garments with seams or cut out the whole joint assembly and test it. For example, I could do a high drop on a whole shoulder, knee, elbow with all the seams that are included. Labs don't do this because they'd need a LOT of whole garments. At least 3, to get 6 samples, and a lot of specific tests. The Darmstadt (AART) can't do these kind of tests at all.

    Thanks for your interest. I hope I can publish some interesting videos for you and other interested riders.
    #13
    Netromsa and Snake Oiler like this.
  14. Jan from Finland

    Jan from Finland Been here awhile

    Joined:
    May 2, 2005
    Oddometer:
    976
    Why not just do the standard test using a heavier striker? It would take into account the impact force.

    Motorcycle crashes are hard to replicate. So many variables. For example, this MotoGP highsider had more than 1 second "flight time". That's close to 3 meter fall height and an impact force of 23.8G.
    https://www.motogp.com/en/news/2020...lease-alex-marquez-highside-crash-data/357536
    #14
    ukAdventurer likes this.
  15. ukAdventurer

    ukAdventurer Long timer

    Joined:
    Jul 9, 2007
    Oddometer:
    1,108
    Location:
    Ulaanbaatar, Mongolia
    TLDR; Just increasing the mass wouldn't be practical or realistic. There's no good, cheap and reliable substitute for gravity.

    However, just adding weight to a joint (let's say the elbow) isn't going to sufficiently make up for the dramatic effect of the kinetic energy from a higher-drop. The researchers decided on 5kg for most tests of impact/abrasion (clothing) and linear impact tests (armour pads). So if we increased that to 7kg but kept the height the same, let's look at the numbers:

    Formula: https://www.calculatorsoup.com/calculators/physics/kinetic.php
    Screenshot 2021-03-06 at 12.28.33 PM.png


    Kinetic energy in joules = 0.5 * m (kg) * v (m/s) * v (m/s)

    10mm (almost on the ground) 5kg @ 0.4 m/s = 0.4 joules (EN17092)
    50mm (near ground) 5kg @ 1 m/s = 2.5 joules (EN13595)

    If we increase the mass to 7kg:
    10mm (almost on the ground) 7kg @ 0.4 m/s = 0.56 joules
    50mm (near ground) 7kg @ 1 m/s = 3.5 joules
    not a big increase in Kinetic energy because the drop height is so low.

    Let's increase the drop height to a more realistic average of 1 metre (which was chosen as the impact protectors test drop height and weight).
    100cm (elbow height) 5kg @ 4.4 m/s (50 joules)

    to equal the Kinetic energy to 50 joules from a low-drop (50mm) we'd need a 100kg weight in the elbow.
    50mm (near ground) 100kg @ 1 m/s (50 joules)

    So maybe you can see why just increasing the mass wouldn't be practical or realistic. There's no good substitute for gravity.

    Let's take a look at Marquez' crash. He landed on his arse so we can't really use 5kg but by looking up some average weights (75kg male) for body parts here we can take 13kg for pelvic mass. Marquez is a lightweight
    (65kg) so let's adjust his arse to 12kg, considering his total mass.

    300cm (extreme high side) 12kg @ 7.5 m/s = 337 Joules
    With a deceleration of 23.5g we can estimate (using this "car crash simulator" his arse weighs 282kg at first impact.

    Screenshot 2021-03-06 at 2.34.09 PM.png
    That's a considerable amount of Kinetic energy, HOWEVER, for abrasion calculations kinetic energy is just part of the equation, we need the kinetic (or static) energy * area to calculate surface pressure - measured in kPa (or PSI). Essentially if we can spread the kinetic energy over time and surface area, we can reduce the abrasive forces. Crashes are complicated. Especially since we are not rigid bodies. We are soft in some areas (arse, stomach, thighs) but very pointy and hard at others (knees, elbows, hips, tip of shoulder). Armour, fat and muscle will all extend the deceleration and mitigate the forces on the suit and body. This is why I want to do tests with all layers, including armour. Nobody is doing this at the moment in a lab, as far as I know.


    Doing some more calculations from known variables, we can see that Marquez's leathers were not in any danger. The forces on the leather were spread over a larger area that's well within even the static load of current testing. His padding, airbag, softness of his arse, all helped him out.

    Marquez estimated butt area at impact 200mm (8 inches) diameter = 31420 mm2 x mass 12kg = 9815g/mm2 / 314 sq inches
    12kg = 26 lbs
    26 lbs spread over 314 sq inches = 0.08 psi = 0.55 kPa (static load)
    611 lbs spread over 314 sq inches = 1.94 psi = 13.3 kPa (impact load)

    En17092 tests with a static load of 18.75 kPa (2.72 psi)
    En 13595 tests with a static load of 25kPa (3.63 psi)

    However, it's never that simple in the real world. Which is why we need to test as closely as possible to a real accident, without using humans as guinea pigs. MotoGP very kindly gives us lots of data but it's at high speeds so it's not repeatable in the typical accidents for the majority of riders. Also, Marques accelerometer (sensor) is most likely in his race hump, so the impact force is measured there. He'd really need sensors in each impact area to get useful data. In his accident, he'd have needed a sensor embedded in his pelvic bone. Not very practical. That's why we have crash test dummies loaded with sensors but even these get damaged frequently. I might get a "rescue mannequin" in the future but a Hybrid III mannequin (or similar) is never going to be within my budget.

    If anybody read to the end, congratulations. I almost put myself to sleep writing this :rofl
    #15
    RedEX, MZ5 and Snake Oiler like this.
  16. ibafran

    ibafran villagidiot

    Joined:
    Apr 16, 2007
    Oddometer:
    1,726
    Location:
    chicagoland
    Long ago, Cycle Magazine rigged a milk crate to swing down from the back of a pick-up truck. Weight was added to the crate to simulate body weight. The underside of the crate had a pad. The test material was secured over the pad with some yellow chalk behind it. Truck was driven along the street and the crate dropped forthwith. A bit of chalk mounted to the crate marked the touchdown point. When the material wore thru, the yellow chalk marked the point. Nice set up for comparative purposes and very 'real world' to me.

    I bet one of you digitally capable FFs can google it up for a link in this thread?

    IIRC, worn jeans out-performed new jeans? Of course, competition thickness leather was the best and kangaroo leather was the best of the leathers. Not all leathers were tested. So bison or wildebeest or gator leather might be better? Whale foreskin leather was not mentioned. No armor was tested. My guess is that cape buffalo horn carved to fit elbows, knees, etc might be the best natural armor?
    fran
    #16
    ukAdventurer likes this.
  17. ibafran

    ibafran villagidiot

    Joined:
    Apr 16, 2007
    Oddometer:
    1,726
    Location:
    chicagoland
    https://www.motorcycleforums.net/threads/abrasion-test-from-cycle-sept-1988.2432/

    I liked this BITD. Very 'real world' to me. Couldn't find the actual article which I remember reading with a lot of glee. The guys had a lot of fun and found out neat stuff. Like old jeans out-performed new jeans. And some seams gave away well before the material abraded thru. That lead to speculation about buying great materials and having a jacket rip apart too soon to provide protection. I had a competition weight leather jacket fail in a crash when the sleeve stitching let go from the cuff to the armpit. Hardly a mark on the sleeve while my arm looked like sausage. Had a leather store sew up the jacket using the original holes but with reeeal good thread.
    fran

    edit: Now that my memory banks have been stirred, I remember the guys moaning about how heavy the 75lb milk crate got after many liftings to fit new test materials.
    #17
    ukAdventurer likes this.
  18. ukAdventurer

    ukAdventurer Long timer

    Joined:
    Jul 9, 2007
    Oddometer:
    1,108
    Location:
    Ulaanbaatar, Mongolia
    I'd like to see the actual video but the description gives me some ideas. I definitely want some kind of visual indication of the end of the test, if necessary. I was thinking of a bright coloured slip liner material which will be different coloured than the main samples. For example: black nylon outer, then aramid, then orange liner, then armour pad. The point of failure would be at the point when orange material is seen - slip liner material has little to no abrasion resistance and typical lasts <0.2 seconds. A GoPro shooting at 240fps will give me start and end times. Coloured chalk could be useful too. Very old school. I like that.

    Here are some tests done with different materials on the Cambridge impact abrasion machine when it was being researched. (Not seeing the same thing with used denim)

    Screenshot 2021-03-06 at 11.54.05 PM.png
    #18
  19. spuh

    spuh Long timer Supporter

    Joined:
    Apr 9, 2012
    Oddometer:
    1,439
    @ukadventurer,thanks for undertaking this project. I think right here on ADVr would be my preferred medium.
    I think the back-of-a-pickup method above was pioneered by Andy Goldfine researching materials for the Aerostich. Since been replicated by magazines. Simple and cheap and the chalk appearing on the road makes an easy to define the point of failure. I like the compact nature of your rig; you could do a lot more test runs with the small samples used. I'm torn between separating the impact from the abrasion phases and combining the two as that's how it happens in the real world. Good on you doing this andI am looking forward to seeing how you progress.
    #19
    ukAdventurer likes this.
  20. ukAdventurer

    ukAdventurer Long timer

    Joined:
    Jul 9, 2007
    Oddometer:
    1,108
    Location:
    Ulaanbaatar, Mongolia
    Thank you! It's not going to work if nobody's interested. I agree that you don't need a fancy rig to test gear. Labs are for convenience and high volume testing.

    Do you remember if the back-of-a-pickup method used a typical drop height - like 1 meter free drop? I've never seen the video. Text descriptions don't mention the impact.

    I'm still working out which would be the best test - luckily I have flexibility to choose. The main question is: is the test relatable, repeatable by me, replicable by others and realistic? My 4 R's of testing. Most labs can only do 2 or 3. Usually they fail on realistic and relatable. But we tend to trust them anyway.

    The chalk idea is a good one - I'd thought about coloured paint in powder form or grease paint (used on crash test dummies) previously but sort of dismissed it as potentially unnecessary. Maybe I'm wrong. I like the idea of being able to measure the distance covered (first impact to failure). GPS only, probably wouldn't be accurate enough. Video evidence should give time from impact to failure.

    Also, colour coded material layers can do this too. For example, if I have black outer textiles or leather, I can use an orange slip liner. The goPro (at 240fps - 4ms frame intervals) should pick up the point when the last layer is reached - we should see orange material flying off. Inner/slip liners would add 0.1 to 0.2 seconds more slide at best. If the outer material is high-viz, I could use a black slip liner material for visual evidence of imminent failure.

    The mobile drop tester uses the same size samples as test labs use (about 15-16cm diameter (roughly 6")) which is nice because typical samples I can request (often for free from manufacturers) are 30x30cm so I can get 4 tests out of that, establishing a mean result.
    #20