New odyssey stampy video

[DontcallmeKenneth]

The sd ones are shadow inceptiv forks.

[Anon.]

They said period.

[streetStreet]

If I remember correctly G mentioned previously that they only tested one pair of forks per competing brand. If this is the case again then it is a bit weird, statistically, to round up all forks for which n=1 to within 0.1 K, and for the one brand for which n≥1 (I assume more than one Odyssey fork was tested) to round up to 650 K+. It's not wrong perse it's just a bit of a weird way to represent your data in my opinion. What is the number of replicates, average number of cycles and standard deviation for Odyssey forks?

It's only BMX.

I'm not the one who made the barplot, just playing devils advocate. For what it's worth I'd never venture elsewhere when it comes to forks, owned 7 Odyssey forks and the fork that came on my complete bike in 11 years of riding. I actually think the stampy test is really cool and testing your product beyond industry standard is the way to go. I just think the barplot is a bit misleading and if I remember correctly G actually mentioned pretty much the same thing previously but called it not being able to publish data of competitors forks due to only one fork per competing brand being tested. So now that they did a second stampy test I'm curious to find out more about the data on Odyssey forks: How many were tested? What was the average number of cycles until failure? Where the results reproducible?

we'd know but odyssey's forks are still on the machinhe.

[G]

G, how much input did you have in the stampy test and the ody forks design?

Lots.

Rad. Just to put it all into context, what sort of real world impact would cause that amount of deflection we're seeing?  Ie 70kg rider from 4ft, etc, etc....

Is the test applying a predetermined load, or is it deflecting the forks a set amount?

It is applying a set load. 5000N for the first half million cycles. All our competitors failed well within this first section, but ours were still going strong so we ramped up the load to 6000N in order to actually break them, and even then it took a further 150k cycles...

It is impossible to draw an exact analogy here with riding. You can ride off a curb with your arms and legs locked and put a huge load on, or you could ride off a 6 foot loading dock and land super smooth. I am working on some strain measurement stuff at the moment to try to get some "examples" of loads applied, but it can never be definitive. I would guess that this (5000N) is about the load that most people apply (momentarily) on their bigger tricks (jumping down a 7-8 stair set perhaps); but this is very much a guess right now.

It looks like this test was the same as the original, cycles of 5000N (at 01:08)

Previous test was 4000N, which was the maximum we could so with a single cylinder. New rig lets us use two.

If I remember correctly G mentioned previously that they only tested one pair of forks per competing brand. If this is the case again then it is a bit weird, statistically, to round up all forks for which n=1 to within 0.1 K, and for the one brand for which n≥1 (I assume more than one Odyssey fork was tested) to round up to 650 K+. It's not wrong perse it's just a bit of a weird way to represent your data in my opinion. What is the number of replicates, average number of cycles and standard deviation for Odyssey forks?

I agree to some extent.

I wanted to present the data slightly differently, but it needed to be a nice clear easy to read visual.

Our fork did 500k cycles at the same load as the other forks (5kN) and just under 150k cycles at a much higher load (6kN). Rather than make this too complex, I think it was fair for us to say that at the same load throughout (5kN) we would easily have surpassed 650k cycles. This is what the bar graph is meant to represent.

Yes we only test one fork from each of our most popular competitors. This is because we have a limited budget, and we purchase each fork in the US and then ship to Taiwan, so we are giving some money to our competitors and spending a lot on shipping and testing.

I hope this all makes sense and seems reasonable, and I am pleased that so many people find this interesting.


G.

[skateparkrider]

I was stoked to see another Stampy test.  But was bummed it was another fork test.  I was hoping for something else, but I understand the reasoning for the second go at forks. 

Also..... this could be a little bit of the Coors Lite talking here.....  but C'MON!!!  Don't you feel responsible in the slightest bit to let people know what BMX products are designed in an inferior/dangerous way?  I know that sometimes you can stir up the shit pot too much to the point where you might fall in and then be covered in shit.....  So I do get it on one hand.  But on the other... at least you can get out of the shit pot, take a shower and not smell like shit after a few days.  Everybody else would be forever smelling of the stench that is doo doo. 

[JFax]

I Think the biggest question is;
Did you use the "Bikeguide, I hold you forever true" code when buying the forks from Empire?

[Bunky]

^^^green box

[G]

I was stoked to see another Stampy test.  But was bummed it was another fork test.  I was hoping for something else, but I understand the reasoning for the second go at forks. 

Also..... this could be a little bit of the Coors Lite talking here.....  but C'MON!!!  Don't you feel responsible in the slightest bit to let people know what BMX products are designed in an inferior/dangerous way?  I know that sometimes you can stir up the shit pot too much to the point where you might fall in and then be covered in shit.....  So I do get it on one hand.  But on the other... at least you can get out of the shit pot, take a shower and not smell like shit after a few days.  Everybody else would be forever smelling of the stench that is doo doo.

Yep, we agonised about this, but BMX is way too small a pond to start taking a big dump in the middle...


G.

[erenes]

The standard "industry" test setup for forks is: +-650N, with force set perpendicular to the fork, <25HZ, 100.000 cycles. Immagine attaching the front wheel to the wall, then pressing and pulling (only with 65kgs), few times a second, 100K cycles. Any entry level mtb 5us$ fork made of cheap steel should pass this test. Make no sense to test bmx forks this way.

Odsy setup looks good. Stampy jig with build resonable headagle, and force direction like real landings. Resonable 500-600kgs force. Love it.

G, it still weird to see so many competition forks snapped clean at the steertube. In real life I see them cracked mostly at the welds e.g. blades/steertube junction or sometimes dropouts. Not too many snapped steertubes around for the last few years. I hope the new forks are not too light/too thin in this critical point.

One more... G, I think I see different frequencies used [0:50-1:10]. I dont think the video is speed up... I'm guessing only. ?

[dude...]

^^^^ yeah it is interesting where they broke. most of the time, ive seen forks break at the welds on the dropouts, or where the legs are welded to the steerer. ironically ive seen a couple of sets of old odyssey forks break off clean through the steerer, although i think it could have been due to a stray ball bearing from the headset rubbing between the steerer and headtube (there was deep grooving).

id like to see a test that puts vertical forces on the fork legs and dropouts as i think this is probably what causes failure the majority of the time from the impact of harsh landings, and right now also from smashing onto the peg for grinds. its gotta be hard to simulate actual riding forces though because theres sideways loads as well and a combination of forces coming from all directions, especially in the case of say an big 360 down some stairs

[G]

The standard "industry" test setup for forks is: +-650N, with force set perpendicular to the fork, <25HZ, 100.000 cycles. Immagine attaching the front wheel to the wall, then pressing and pulling (only with 65kgs), few times a second, 100K cycles. Any entry level mtb 5us$ fork made of cheap steel should pass this test. Make no sense to test bmx forks this way.

Odsy setup looks good. Stampy jig with build resonable headagle, and force direction like real landings. Resonable 500-600kgs force. Love it.

G, it still weird to see so many competition forks snapped clean at the steertube. In real life I see them cracked mostly at the welds e.g. blades/steertube junction or sometimes dropouts. Not too many snapped steertubes around for the last few years. I hope the new forks are not too light/too thin in this critical point.

One more... G, I think I see different frequencies used [0:50-1:10]. I dont think the video is speed up... I'm guessing only. ?

The industry standard tests biggest problem is that they dont mount the fork in a headset, they clamp the steerer directly. This is absolutely ridiculous and makes the test almost completely worthless. Since we started testing this way several MTB companies have followed suit and now test using a proper headset mounting.

Dropout weld cracks are mostly due to peg use and landings off straight, although this may be a fairly common mode of failure it shouldn't be. We very rarely get forks back on warranty with this issue unless there is obvious evidence of peg use with a crappy hub axle (ie. heavy indenting and deformation of the inside of the dropout) the refinements to the 41 thermal process welding and the greater use of big female axles have fixed this for us.

The base of the steerer is where I always expected to see failure. This is where there is maximum bending moment and the unavoidable stress concentration of the headset seat, but different loading regimes will always have the potential to change the mode of failure. If a fork "sees" more side and torsional loads then weld failure is possibly more likely.

Yes in an ideal world we would test multiple scenarios, but we have to deal with the realities of the cost of testing.


G.

[dude...]

theres a few people around here still running 14mm front wheels, but the supply of 14mm dirt forks to go with are drying up as they all end up cracking above the dropout on the regular peg side.
not to say this is a fault of the forks, cos everyone still running 14mm forks is doing so because theyre pretty harsh on them. noone bothers warrentying them because most of the time theyre 3-4th hand anyway and thus the warrenty is void, plus theyre not gonna get a replacement set of 14mm forks even if they were able to warrenty them anyway.

ive personally cracked some dirts like that, i think the 06 versions, while using a 14mm odyssey hazard hub

but yeah safety wise its probably better to make sure they dont just snap off clean at the steerer suddenly, if the dropout snaps at least the whole front end isnt necessarily gonna fail and put you on your face. and yeah i can understand its already expensive doing the stampy as is without doing additional testing

[JFax]

Is there any difference in strength between fork dropouts made for 3/8īs and those made for 14 mm? or is the difference neglectable?

[dude...]

dirt forks had slightly thicker dropouts than the race forks. wasnt really about the fact that they were 14mm, only that people were cracking them then struggling to find replacements. everytime dirt forks come up secondhand they get snapped up really quickly

[erenes]

(...)
The industry standard tests biggest problem is that they dont mount the fork in a headset, they clamp the steerer directly. This is absolutely ridiculous and makes the test almost completely worthless. Since we started testing this way several MTB companies have followed suit and now test using a proper headset mounting.
(...)

I would say there is one factor that is about 100x times more ridiculous than 'unreal industry standards'

It is.... to allow outsourcing R&D guys, engineers, agents, official EN testing center engineers - seting up your babies, your products, allowing them to install products into jigs before starting the test. Most of them never rode a bike. If they rode a bike, they never jumped. If they jumped, than maybe the curb, they never rode real BMX.

Of course you can prepare detailed guides about setting up position, angle, tightening torque for your tested products, sketches, drawings, video-guides, anything you want. But there are 10x more setups and cases possible with testing machine elements / jigs than with your products itself. Comparing bars of different width would you setup up them with same width or same distance from the bar ends? Both could be ok depends what you plan to compare/test.

Then easier one, but not always - setting up the machine program itself. I saw so many screw-ups.

I saw so many RIDICULOUS testing done. E.g. comparison fatigue test of saddle+seatpost set of two brands. But the setup, jigs used, and especially the way they bolted saddle + post togeother was so ridiculous, screwed up, upside down. You could immediatelly notice that certified testing center personal never rode ANY bike. They were under influence of HEAVY DRUGS, Im sure.

I think that if you are not there (in Taiwan or somewhere) playing with your babies and competiton products - the test does not count! Only testing done 'inhouse' make sens in long-term.

G: any comments on the different frequence ?