Try reading what i actually write. What you are doing now is basically just setting up a bunch of strawmen.

---

http://bikeblather.blogspot.com/

here is what you said:

"a small bit of time with an instrumented stem and a data acquistion system would get you all you need to know about the possible loadings to create an effective design. Apply a reasonable factor of safety, and the engineering is fairly routine...especially in this day and age of PC based analysis tools... This is true of basically the entire bicycle as well."

your view is that testing is only a validation of engineering, and that testing failures should not be inputs into the design process. you singled out cervelo as a company that operates in the way you describe. i would be shocked if damon rinard or any of cervelo's engineers agrees with you that failures in testing should not be taken into consideration during the design process.

the reason we have oversized bottom bearings in fork crowns can be traced back to repeated failures in testing just above the crown, confounding the engineering. you think this design paradigm is "tinkering", which i interpret as an amateurish approach to product design. cervelo has a little experience in fork and stem failures - as do trek and, well, most companies that have been making these products a long time - and i am quite confident these bike companies consider such failures design inputs.

if bikes enjoyed a stasis in design, yes, you would be right, most of what we need to know about the forces applied to a bike would be knowable and we'd design with ultimate confidence. but that's not the case. the minute you change a hub flange design, or use straight pull spokes instead of elbow-style spokes, or change the epoxy that holds 2 things together, or put bosses in the top of your top tube, or a hydration system inside your down tube, you introduce unknown unknowns.

i know one fork maker, early in the carbon fork days, who had it all figured out. it engineered its fork expertly, figuring out all the loads, all the stresses, and it tested its forks to make sure they would pass, and they passed. and then came the failures in the field, because there was one force this company forgot to consider: what happens during braking. the engineers just didn't think about this. they forgot. it never occurred to them. that's why you test. it's not just to validate your superb engineering, so that you can high five everybody when you do your victory lap. it's because sometimes you aren't as smart as you thought you were. and when you have that failure, you're a fool if that failure does not inform the changes you must make in your design.

---

Dan Empfield
aka Slowman

Nope. Never said that. In fact, if you look above you'll see where I said that testing failures feed back into determining what was missed in the modeling: "..if the results don't agree with the modeling, you need to determine what was missing or incorrect in the model so that the models can improve for future analysis." Validation is of the modeling as well as the actual part.

Slowman wrote:

i know one fork maker, early in the carbon fork days, who had it all figured out. it engineered its fork expertly, figuring out all the loads, all the stresses, and it tested its forks to make sure they would pass, and they passed. and then came the failures in the field, because there was one force this company forgot to consider: what happens during braking. the engineers just didn't think about this. they forgot. it never occurred to them.

Gee...I wonder if they'd instrumented up a fork and ridden it around awhile during the load definition phase if they might have caught that braking load ommission? :-/



Never said that either. Of course any testing failures inform changes in the design. But, they also inform changes in the modeling so that failures are less and less likely in future designs.

---

http://bikeblather.blogspot.com/

Many (most) mil-aero OEMs don't allow 7075-T6 for use in structural applications (anything other than sheet form). We'd generally opt for 7050.

"Nope. Never said that."

maybe i just misunderstood you when you said that testing should be for validation, "NOT as design tools". can you again apologize if i misunderstood? ;-)

i'm not sure, but i think i can narrow the difference between our views to how it is engineers respond to failures. i think your view is that even if the engineering suggests a part will not fail (let's stay the area of a stem's attachment to the steerer), if it fails in testing you go back and see what you did wrong in your engineering. that is the only solution - find your engineering problem, find your modeling problem, remodel the part, retest the part. if it breaks again, find your modeling problem, rinse, lather, repeat. i'm saying that when you're on the razor's edge of safety and performance, you might find that there's a gap you can't bridge because there's something that happens that you can't anticipate.

i will concede you this: it's much easier to match engineering with testing, if you make your own testing tools (or you understand how somebody else made your testing tool) and your testing tool is fairly rudimentary. it's much more likely that you're going to have this gap between real life and life explained by engineering in the field, because there's a gap between what we think will happen to a bike (or a bike part) and what does happen to it (or, if the testing apparatus you're using is artfully built, very complex, and very good at replicating what happens in the field).

this is one of the problems in our industry. i had a specific part failure at merlin, a machined seat stay part breaking in the field, our industry standard test machine could only break a chain stay, we never broke a chain stay in the field, just a bunch of machined triple clamps (turning 2 seat stays into a monostay). the test machine never broke a triple clamp. you can't engineer properly what you don't yet understand and when that's the case, it's the failures that inform the design process directly, bypassing the engineering.

---

Dan Empfield
aka Slowman

Can you explain why? I have no experience in mil-aero.... but in my industry we use 7075-T6 a decent amount.

"Uncle"

---

http://bikeblather.blogspot.com/

Stress corrosion cracking, you can read a bit of the background here, but every mil-aero prime I've worked with prohibits 7075 in anything >=0.25" thickness (i.e. other than sheet). 2024 is also not allowed.

Thanks for the link. Learned something today. That also explains why we don't mind using it in the industry I am in...low service life and much shorter times between regular cleaning and inspecting parts.

you're calling uncle? or you think i'm calling uncle?

---

Dan Empfield
aka Slowman

I called Uncle. You obviously don't understand the role of measurement in the design process. I'm tired of repeating myself...

---

http://bikeblather.blogspot.com/

"You obviously don't understand the role of measurement in the design process."

in the last 15 years of the 20th century, quite a few people entered the bike business who believed as you believe today, that, as you said, "the engineering is fairly routine...this is true of basically the entire bicycle as well."

these people were far more dangerous to end users than were the "tinkerers", because the tools for engineers back then weren't nearly as advanced as they are today, and because a century of tinkering had given us bikes that did not break. until 1985 or so. that's when weight really became an imperative. we had a lot of bike and bike part failures in the 90s because the engineering was incomplete, and the engineers built testing machines that matched their knowledge of stresses on a bike - which was incomplete.

today, engineering is an imperative in bike design because the weights are really down to the nub. you can't be a tinkerer anymore. (why you insist on implying that i am anti-engineering is god's own mystery. i said the opposite a number of times.)

today it's the opposite. the tinkerers are now a bigger hazard than those in the industry who think "engineering [for 'the entire bicycle'] is fairly routine", when you're talking about forks, stems, cranks, handlebars. but if you look at the failures in these parts, they are as likely, or more than likely, to come from companies who employ top rate engineers. why? because the engineering is not routine.

we are now in the state of perpetual recall. we are always in the middle of a recall from a major company employing expert engineers and big testing budgets. the very best engineers i know in this business are those who blend excellence in their craft with the humility that comes from acknowledging that engineering for bikes is not fairly routine.

---

Dan Empfield
aka Slowman

For the love of god, all you "engineers" keep stating 7075-T6. Really? How thin is this stem going to be?

7075-T651. Block material, not sheet metal.

It is the core of all aerospace. The majority of parts, including structure, are made of aluminum. 7075 being the most often used.

We are a machine shop and we make parts for the aerospace industry.

But if you're going to throw around specs on material, the correct material is 7075-T651.

I remember those posts, but I missed out on it. Have any to sell?

I could have been pedantic with the OP too, but its pretty clear that I understand the forms when I specifically addressed them. On an actual drawing, 7075-T651 would not be good enough either, you'd also need the QQ or AMS spec and the material basis. There's also no such thing as block material, as you know.

All of which is besides the point. 7075-T6XXX is prohibited for use (in the procurement specs) in anything but sheet form in structural applications or applications that require corrosion resistance, on most platforms, including both the military and commercial arms of Boeing, LM, NG, Airbus and GA.

7075 has been largely replaced by 7050, although I think that its more of a PIA to machine... There are of course exceptions, and a deviation would need to be sought for each such case.

So, in the case of a bicycle stem, that is a structural application, that is covered in sweat, there are better choices.

For vertical lift, we seem to use a lot of it. Of course, most our parts are not exposed to the elements and are internally based.

I'm pretty sure Cnc machined 7075-t651 with a plating and paint would be more than adequate for a bike stem. Heavy as hell in comparison to forged or die, but pretty bomb proof.

And we still calm it blocks when we procure material. Call us barbaric.

Just trying to match snark :)

The world is ending, an engineer with a personality....

Noted!!!

Regarding the material choice, we are using T6-7075 because of its excellent mechanical properties, and also because it has far exceeded all of our testing protocols (and easily surpassed both the CEN and CPSC standards tests as well). All of our stems are ALSO coated with an electrolysis process (a type of anodizing) that adds additional corrosion-resistance to the part. We are confident that there will be ZERO failures with the new stems, period.

In my experience there are ALWAYS unknowns in real-world engineering, despite the best efforts of the engineer.

I agree with Tom that, theoretically, the load cases and engineering parameters of any part can be accurately and completely identified. But I also agree with Dan that even the best engineers DO NOT identify them all, even with the highest levels of knowledge, experience, and diligence. There will always be a gap between practice and perfection. But TriRig is committed to closing that gap with all of our efforts.

---

--
TriRig.com

Only other quality check I can think of is to have each piece sent for MPI or FPI to check for micro abrasions. The anodize should be fine for corrosion. A quick paint and poly coat would seal the deal. Powder coat would be a nice touch!!

If I were to make this part, I put the cost per piece based on a qty of 100pcs around $200. Material is inexpensive and easy to get. One time charge for 5 axis machining (to reduce fixtures) Cnc programming, cmm inspection programming, and fixtures, are pretty expensive but can be spread over the 100pc lot. The machine time is expensive too as you will likely set this up in a 5 axis and make 1-4 pcs at a time. Gets more expensive if put in 3 axis machine with more fixtures. Deburring and buffing or sand blasting or bead blast will be a lot charge. Mag or fluorescent inspect, anodize, prime, will be a per piece price based on size.

This is not a cheap fix for you nick, nor is it easy to raise the quality flag in yourself. I didn't take liberties at condemning you in the other thread regarding your customer service as I've had mixed emotions as a consumer of many omegas and an alpha. But this call to action trumps any disappointment I've had in past communications. I applaud your efforts on this.

Now, please double check the stress on the alpha, I love that damn bar and I really don't want to see that pop up on a thread like this!!! No way I want that in a Cnc milled aluminum version.

Good job nick. And good job dan for letting nick come on the forums again to post the info.

I'm looking for some input on what folks are using as an interim replacement .. are there any stems that use the same amount of steerer tube as the Sigma? If not what have folks rigged up (no pun intended)?

The syntace flatforce looks close in height.

hmm... I wonder if the flatforce would work based on the spec's below? I think the Sigma uses 22mm of steerer height?

It may be moot as the flatforce is a MTB stem that appears to come mostly in very stubby / short lengths and does not appear to be widely available either?

Steerer Height 27 mm (55 mm) Minimum Insert (steerer) 22 mm

It comes in lengths up to 111mm(longer than the sigma) and appears to available through QBP(i.e. any bike dealer).

Anyways, a person with a slammed sigma may have too short of a steerer but if a person had a spacer of a few mm, it should work, I think.

Here is my solution to the Sigma problem:



I got a friend who has a lathe to make the two pieces for me. I must stress that I have not used it yet, and anyone who copies this design does so at their own risk. But I am posting it immediately because I know some people will be really struggling for an immediate solution and may want to get on with having something similar made as soon as possible.

The piece on the left is intended to be bonded into the steerer. The grooves on the lower outer surface are to improve the ability of adhesive to retain it in the steerer. It has an M6 thread that goes all the way through the centre, and a larger female thread at the top to allow the piece on the right to be screwed into it. The idea is that the piece on the left will be used on its own when the Sigma replacement is received.

The piece on the right also has an M6 thread all the way through, and a larger male thread protruding from the bottom to allow it to be attached to the piece on the left. With this in place, a conventional stem can be used until the replacement Sigma is received. Note that it is only intended to support a stem where the steerer comes to around halfway up the stem clamp, it is not intended to have a stem clamped entirely or mostly on the extender. As I said, if you want to get something similar made and use it yourself, you do so at your own risk.

The friend who made this for me doesn't want to make production quantities for other people, so you're on your own for getting someone to make something similar for you.

The lower part could also be made using an expanding bung similar to existing commercial products, but it would have been much more complex for my friend to make, and I have a TT tomorrow that I want to use this for, so simplicity won. I'm happy enough to have it bonded in, as together with the extending piece, it covers all my expected use cases. The extending section protrudes far enough above the stem that a HED lollipop can be attached to replace the Sigma's bottle mount.

The total weight is 96g, and it's made from this alloy:
http://www.aalco.co.uk/...6T651-Plate_148.ashx
I'm not specifically recommending that, it's just what he already had, but it seems reasonably suitable, with good strength and corrosion resistance.