rpenmanparker 

Your points are well taken, but you don't consider how a wheel built to optimize radial NDS spokes (or even DS) would compare. What would be the effect of putting all the radial NDS spokes heads out to get the tensions closer on the two sides? What would be the effect of putting all the 2X spokes heads in on the DS to further bring the tensions on the two sides together. Would the wheel with 1/2 radial spoking then be as good as or even better than the wheel designs tested. Dunno, but that is what more test could show. Until reading this article, everything I have read about best wheel design has been unsupported by data. I don't mean the data didn't exist, but that it wasn't provided. This is a watershed event for the technically savvy consumer. More information will provide even more knowledge about "best" wheel design or even how to make any design of comparable durability to others. It is too early to draw conclusions as so much is still yet to be done.

Jiggle

I think wheel builders already know what the "best" wheel design is. But what will make them stand out in the market? (DS radial spokes, paired spokes, gimmicky aluminum spokes, etc...) What lacing is easiest to train a monkey to build? (NDS radial), etc....

rpenmanparker 

Most, like Psimet, know of what they have tried, what has worked best for them. That is not the same as knowing the absolutely optimum design (or the most recent approximation of it) according to finite element analysis.

qcpmsame 

It really bothers me that I read every reply, understood what was said and enjoyed the give and take in this thread. I have been told I have a very bad case of Engineer's Disease and I am a member of the Hairy Eared Engineers Society, too. An FEA discussion that ENTERTAINED is a one off for sure.

Bill

MikeyBoyAz

Something I wonder about all this analysis is the inclusion of asymmetric rims where they push the spoke hole far left to reduce the difference in spoke tension. Shimano and Mavic both seem to like doing it on their shallow alloy rims.

pro's thoughts?

rpenmanparker 

Not a pro, but I do have experience with the asymmetric rims. They work great to help balance the tension side-to-side. If that would reduce some of the extremes measured by the testers, then it would likely be a good thing for wheel performance and durability. They likely reduce wheel stiffness a bit due to reducing the NDS bracing angle, but I don't think the effect is significant.

Psimet2001 

Meh.

They can help balance tension a LITTLE bit more, but the structure then becomes asymmetric. In reality the rim gives the wheel the vast majority of it's stiffness and response. The spokes - in essence - are there to adequately support the rim structure. With an asymmetric rim structure you get different qualities depending on the direction you are running.

If the basic rim structure is sound to begin with - good alloy, appropriate area moment of inertia - then there is nothing wrong with asymmetric. The net is a slightly better tension balanced wheel. Yay. The drawback is when the rim structure really isn't sound to begin with.

Prime example - (forgive me Velocity - you know I love you) - Velocity has had some serious extrusion issues since relocating their production facility to the US. The extrusions are done outside so this is a supplier issue. The issues have hit the A23 HARD off and on for the last couple of years. When the metal is good the A23 makes a good wheel. When it's good the A23OC makes a great disc option for cross in the front and rear. When the metal is off....they both suck....but...the OC sucks more....lots more. I've got a short stack of them that look more like an attempt at an art sculpture than a rim.

Can you stack that to the metal alone? Maybe. Maybe these were just worse than even the bad lots of Symmetric. Thing is they both felt the same, but the compromised side of the OC just felt like a piece of tin foil.

YMMV, but in this instance I'll take a symmetrical A23 any day over an OC.

.....actually I'll take a long list of rims before I take an A23 ever again, but....you get what I'm saying.

Psimet2001 

Now - a general Engineering statement.

Most of you know I am one. I like to think of myself as a slightly rarer version that is made up of half Engineer (dad) and half Artist (mother). In practice it means I can get analysis paralysis as the engineering side of me approaches issues with the mentality of "there HAS to be A right answer"! 90% of the threads in this forum can be classified as such "I'm here...this is my problem...what is THE SINGULAR right answer?"

Luckily the artist side of me steps in and I start looking at the big picture so to speak. I have to.

A long time ago I learned that any and all analysis ever done is heavily biased. So much so I tend to spend more time looking for the bias than ever doing anything with the result. Financial, technical, doesn't matter. Numbers are only "good" for what the person specifically looked at.

....and yet you can't go through life saying that all analysis is impossible. It's a constant struggle.

FEA is no different. Those who have actually taken classes, theory, etc regarding FEA understand this. It's more similar to a craft in application rather than a science. There are no hard and fast rules. Experience and intuition has to be the guide in determining the proper element creation, boundary conditions, loading, etc. Unfortunately just about every good 3D modeling software package has a built in FEA module. In essence most people that can draft in 3 dimensions can now type in some variables and get an analysis. GIGO....Garbage In: Garbage Out.

Have I ever wanted to do a full in depth FEA of wheel assemblies? Absolutely. On my "list". Will my results be "published on my site"? No way. I would do it solely for the purpose of building better wheel assemblies. If I am confident in my selection of variables then I will only end up with even better performing/lasting wheels....something I've already gone to great lengths to achieve. The wheels speak for themselves. I don't need to publish an FEA analysis - an analysis that is too complex for the vast majority of customers to make use out of and too simplistic for those that do understand it. It's like publishing FEA analysis of crash results for cars....or just publishing a cool photo of the model and a crash performance rating...

Like I said before - the general conclusions listed tend to support, by trend, the same results I have seen in other direct studies on lacing effects on wheel stiffness. All show that x lacing always out performs radial lacing, and that the higher the X the marginally stiffer the system becomes. Anecdotal evidence over thousands of real world wheels built directly supports that. So....big picture - is X lacing better than asymmetrical lacing? Yes.... Across all rims, hubs, drillings, loading applications, symmetries, etc? Yes. Why would anyone ever do anything different?

Have you ever seen the rants people can get into about lacings? I get tons of inquiries where customers tend to tell me that asymmetrical lacing is better. When I ask why they think that they basically said because the internet told them so. Why did the big guys start doing it? It wasn't from a marketing standpoint as implied earlier. Simply a different engineer working with a different bias stated his/her case for why it was better. It most likely sounded good. People did it. In the end....it doesn't actually make enough of an appreciable difference for anyone to ever sit there and point and say "HEY....that way works better". That's why people still debate it....because it doesn't really make a difference. If it did it wouldn't be debatable.

At the end of the day if the differences are small and Asym LOOKS better, performs even remotely better aero wise, or drops the weight of the wheel by 2-5 grams then the big guys are going to do it. Volume runs their business and it has way more to do with moving the units in the supply chain than any other factor in the wheel experience.

rpenmanparker 

My experience with Aerohead OCs goes back several years, Australian production I imagine. But I agree with you completely. I have been off Velocity since I started having trouble getting them true without areas of very high tension due to unevenness in the extrusion or lack of initial flatness. Whatever. I just stopped using Velocity at all. I liked the side-to-side balance improvement I got with the OC, and I don't know anybody else offering that level of offset to the do-it-yourselfer. But I am just making do with symmetrical rims now, because I don't want to use Velocity.

ovoleg

I can hit 1450W @170lbs, this study is flawed. There are guys in my area that can easy do over 1500W. What the hell kind of extreme is 1100, thats weaksauce.

ovoleg

alot of truth here. FEA, depending on your boundary conditions could be anywhere from
A) junk results and pretty graphs
B) only modeling half the picture
C) legitimate analysis

The hardest part about FEA is not generating the pretty graphs or making the models its identifying the boundary conditions. Figuring out how to apply the load and where and where to affix the component is 99.999% of the battle. Unfortunately, people make asinine assumptions most of the time and you get ******** results that everyone accepts as fact.

I don't know much about wheel building and I skimmed the article but it looks like a paper I would write for one of my engineering classes.

<--mech engineer.

canam73

You should review posts 15 & 16.

canam73

In case you are interested, IRD offers the Cadence VSR. It is basically a Kinlin XR19 with 3mm of offset.

MikeyBoyAz

mine are the DT Swiss RR440A (rear) I have never measured how offset they are, but it's close to vertical on the NDS.

rpenmanparker 

Last time I checked IRD rims, the offset was advertised to be much smaller. At least that's how I remember it. Did they make a change that you know of or is this a newer model? Thanks for the information. Kinlin is my current favorite, and I do appreciate eyelets, so the IRD would be a good choice if the offset is as large as you say.

rpenmanparker 

Rob, if it IS possible to lace X patterns with all the heads facing in or out on each side (as you mentioned earlier), why do you not do this when working with the hubs that have the largest side to side differential? Heads out on the NDS and in on the DS with either 2X or 3X most commonly. Once I get below 50% of DS tension on the NDS, I get a little nervous. Why not use this "trick" to get back above 50% or even a little higher on those "worst case" hubs without having to resort to assymetric rims? Does the low NDS tension not bother you as much as it does me? Is it really no issue? Thanks.

canam73

You could be right, I pulled that out of an old forum post. Although it was from Mike Garcia.

rpenmanparker 

Took me a while to find it, but I finally did. Here is a page that shows the IRD offset to be only 1.5 mm. I don't know the validity of this publication, but taht rings a bell regarding what I remembered from previously investigating this question. The Velocity Aerohead OC has 4.0 mm offset, significantly more.

https://leonard.io/edd/rim/361-ird-ca...metry-rim-700c

canam73

Well, even 1.5mm would bring a Bitex hub back in line with everyone else. If that's what you are using.

rpenmanparker 

How'd ya guess? Actually I would like to bring Bitex more in line with American Classic, but that ain't gonna happen. 4 mm would get me a lot closer, however. No matter, I like the idea of the IRDs and may invest in a pair sometime soon.

canam73

I guessed because the only reason I know of the IRD rim is in regards to building with BHS hubs. I ended up going to novatec, mainly for the rear geometry.

canam73

While we were discussing this yesterday I took the liberty of emailing IRD on the offset. Per Evan at irdinfo@interlocracing.com it is "about 3mm".

Road Fan

No, you aren't! The researcher presumably started with some particular research goals that didn't encompass all the particular questions folks here might want to see explored. That doesn't mean the study is flawed, it means the study should be evaluated within its goals, and those goals should be clearly stated. I'm not sure if they were clearly stated.

The research might be part of a bigger research program that might include other technology and configuration options. At the end of the day STEM is funded by Williams, and under these conditions Williams generally owns the results. And the owner can decide what results can be shared outside the company and what should not be shared. If Williams is looking to evaluate new design options, I'd think they will not share all of that. If you want to see what they found to be a design improvement, watch their product line-up.

rpenmanparker 

Thanks. I am really glad to know that. Their drawing suggest that, but the only available specification information says 1.5.

rpenmanparker 

Sound reasoning. I agree that study is not flawed, maybe just the first of a series. And you're right, we might not see all the results due to their value to Williams.