BikeTechReview.com

Since the super aero Hooker brakes are nearly unobtainable, the latest design by Matt Simkins looks the goods. Claimed to save 3 Watts over regular dual pivot brakes...

Discussed on Slowtwitch.

Image -

http://i.ebayimg.com/14/!BeoZp8!&#...zfo3NQ~~_12.JPG

With the UCI and their rule change to include items such as bars and seat posts to conform to the 3:1 rule, could brake designs like these be impacted?

no way am I going to make it through a 10+ page thread over on that site...so, I googled simkins:

http://www.simkinsdesigns.com/

and the site seems lacking in supporting any aero benefit claims as of 11/13/09.

That's awesome that he has designed this brake and is giving things a go! That's really cool to see. I wish him the best of luck.

Eyeballing the design makes me wonder how much he left on the table simply due to CAD modeling limitations.

What wind tunnel did he use to support the 3w (at what speed) claim?

3 Watts appears to be attributed to Tom Anhalt field testing the front brake only (page 10). Tom suggests a more detailed write up will be going onto the Simkin designs website shortly. Having just checked the Simkin's website, I hope at the same time Matt fixes the typo's and neatens up his website...

More comments from the thread - (Matt Simkins Page 6)



(Tom Anhalt Page 6) Answering which centre pull it was compared to...



Computer modelling has been used looking at the image from his site.- http://www.simkinsdesigns.com/egg_drag_testing.htm . I don't understand the image so can't say whether anything valuable can be gained from it.

field testing and 3 watt claims - huh - I wouldn't be comfortable with that based on my own personal experience using a halfpipe, powertap/srm and doing lots of laps.

Can't wait to read the detailed field testing protocol description so I can try my hand at it.

As for the other stuff you posted, not sure what to make of the commentary.

The brake does seem to reduce frontal area, a bit, though, eh?

Out of curiosity, what is the total frontal area for a dual caliper brake anyway?

Honestly not too sure of how much benefit really can be had from this brakeset, but when I was told about the brakes and topic I thought it might be of interest here. A couple of guys on other forums have sets on the way, so I look forward to more reports from them on what they think...

On the Simkin page I linked, he does say that if you want the full details he will email them through. So I might just do that.

Without having the two different brakes side by side, its hard to know the actual size of these brakes against the usual dual pivot calipers. Hopefully someone will post comparison shots shortly.

Dalai wrote:


I definitely appreciate the heads up - I can't remember seeing this before.



Let us know if he comes through with the info.



here's an estimate of the frontal area on a front caliper:

pad arms: .010 x .030 =.0003 m^2
body: .020 x .070 = .0014 m^2
upper arm: .02 x .030 = .0006 m^2
total=.0026 m^2 -> at 30mph at sea level and assuming it's shaped like a flat plate...that'd be 0.5*1.2*1.1*.0026*13.41^3 ~ 4 watts.

CxA for those area estimates (someone else feel free to chime in with a better estimate -> I'm just pulling those #'s off the top of my head) would be 0.0028 m^2.

So, that's some crude math to ballpark the effect of a bare caliper in the freestream all by itself - no interaction with anything, and ignoring any potential cable routing/exposure effects.

I don't think frontal area is that relevant seeing the fork is there anyway. Any gains would be in Cx, interaction with the fork.

It does look cool but I am not sure how much saving in total drag is possible.

It would be hard to top the new Trek setup however.

Dalai wrote:


I did send him an email, and Matt sent me a multi page test that describes the field testing done. He chose not to send me the results from his homemade wind tunnel testing, nor the results from his "computer models". The document that Matt wrote and sent me is well constructed and written. If I were advising Matt, I'd suggest that he put the same effort into his website that he put into writing that document that was sent to me. The website is probably costing him sales if it continues to look like it did on 11/13/09.

The field testing method and venue are nearly the same as I've detailed/attempted previously in various threads throughout the BTR forum (venue was a "halfpipe") ...so, it doesn't appear to be anything different than has been discussed before here at BTR...even the error estimation method employed was originally discussed here.

The claim of 3 watts is at 28 mph, it seems, so, I'll have to revise my estimates above accordingly.

There was one repeat run done for the dual caliper brake and it was 1.8 watts different than the original, same configuration run. There were no repeat runs done with the simkins brake.

A cursory "eyeball" examination of the plots provided in the report indicates that conditions might not have been consistent between the stipulated "good" runs...access to the raw files would let others confirm or refute what I am seeing. Hopefully, the raw files will be shared.

That's all I've got - forgot to ask if I could post the pdf that Matt wrote and sent me, but he had a very fast response to my original email, so I imagine he is eager to share the results and his perspective.

kraig wrote:


and now revising this estimate based on the new information that the 3w claim is at 28 mph

at 28mph and rho = 1.228 (what the 3w # is based on) and assuming it's shaped like a flat plate...that'd be 0.5*1.228*1.1*.0026*12.52^3 ~ 3.4 watts

Just to be clear, if one were to completely remove the front brake (or make the estimated brake "invisible to the wind" with a Cx of 0.0), one might theoretically save 3.4 watts.

Anybody have a better estimate/approach to what I'm trying to do/quantify here?

whodesigns wrote:



yeah, I hear what you are saying...having seen the field test configuration images (man, I wish I'd asked matt if it was OK to share the pdf he wrote and sent me) I think there is a little bit of a frontal area difference - primarily in the quick release/upper arm barrel adjuster holder area. Seriously, if anyone is at all curious, I'd suggest sending Matt an email and reading through the report/pdf that he wrote and will eventually publish.

Shaping wise...I don't know, the simkins brake "feels" like a stealth fighter in it's shaping. Not super organic/efficient to me (hence, I wonder if it's being limited by modeling capabilities), so, I'm not sure how much is there from an "eyeballed" shaping perspective. Maybe a skooch, but, there was probably some gains left on the table from a Cx perspective.

I reckon only a beta sweep at the tunnel here in San Diego with the pedaling speedplay dummy would convince me of any real potential gains to be had! )

Hey, speaking of Cx (and "A"), in it's current embodiment, it sure feels like the simkins brake might even be worse than a standard brake at typical betas (y'know not equal to zero, or e.g - typical triathlete speeds doing IMAZ/kona conditions). Anybody have any thoughts on this?

Maybe the bulbous carbon fiber faired Zipp contraption thingy that reared it's head many moons ago really was the real deal??? )

attached is Matt's article that he wrote (posted with his permission).

enjoy!

Any pics of the brake actually mounted on a fork? ...and with cable routing?

@SteveP - photo's are in the attached pdf.

@whodesigns - If I get a set, I will lend them to you to test.

I've attached the pt data that matt provided me - I've converted them back to their original csv format. The assumptions you'll need to make sense of these files are in the pdf previously posted in this thread.

My own personal takeaway when taking a cursory look at these data (and placing the fact that these are field test acquired data into context) remains consistent with what I've previously stated:

1) conditions seem to have changed across runs
2) field testing is difficult business
3) one could probably convince themselves that they are more precise than they actually are when using these types of data reduction techniques.

enjoy, and I encourage those of you that can correct for data drops/repeats to look at these files, as it appears there are some present.

Please let us know what you find after examining these files!

Well, since no one else has taken a look at these files, I've taken a few seconds to just run the raw data (of the claimed "good" runs) through my canned script:

[attachment=0:1c7gyox2]<!-- ia0 file_compare.png<!-- ia0 [/attachment:1c7gyox2]

To me, the power data provided by Matt (as visualized above) suggest that more trials were needed in order to add meaningful insight into the question "is the simkins brake more aero than the baseline configuration tested?".

Anybody seen any more images of these brakes? I'm curious to see what the backside of the frontbrake looks like, and also the what the "front" side of the rear brake looks like.

I saw a pic of the back of the front brake on Slowtwitch. It looks like it is 'skirted', where the outer 'shell' (hence why it is called the egg brake) comes all the way back, almost to the fork. I still think that it would be cool to make a bayonet style front end with a 1" steerer fork with a rear mounted center pull brake calipers which go through the head tube of the bicycle (which is no longer filled with the steerer of the fork). In this way, you would get a really clean mounting, and have the calipers in the wake of the fork crown. If you had an argos fork, you could even construct cantilever pods on the back of the fork and construct some traling edge shaped composite cantilevers...if only I had time and money..........I would probably still be slow and wrong.

Stephen J

kraig wrote:

This is all I have so far:
[url:3ljfu6yf]palladinospics.shutterfly.com/27[/url]
[url:3ljfu6yf]palladinospics.shutterfly.com/28[/url]
[url:3ljfu6yf]palladinospics.shutterfly.com/29[/url]
On a P4.
Let me know if you want another angle.

Steve,
Would you mind if I used one of your pics for the simkinsdesigns.com site? A P4 will always make a bike part look good.

Regarding testing, real world testing using the Chung Method is pretty tough in any event. Tom did it in the desert so he actually got really good data. Additionally, Tom has done that testing many times in the past and is an expert at it. I have talked to folk in the Bay Area and there are very few places suitable for that kind of testing. Traffic, uneven roads, or wind will really screw things up. There is no way that I could have gotten data that nice myself.

Regarding Kraig's comment on the stealth bomber look, the CAD software these days actually is phenominal and can do organic shapes very well. The real constraint ended up being the machining. CNC machining allows you to get really thin walled structure, deep sections, and allows you to use much stronger materials than can be used for forging. The only drawback (other than being expensive) is that the machining time is constrained to the number of required tooling paths. In other words, the more times you have to swipe the part with a mill, the longer it takes. To get a perfectly round part like a ball, or an egg, you would theorhetically need an infinite number of tooling paths. The Egg Brake CAD model actually consists of hundreds of tooling paths and thousands of arcs as it is.

By way of comparion,



are all pretty angular (non-aero) examples of machined parts.

It was great looking at folks analysis on the writeup. I am going to work on a paper regarding computer modeling that folks should really dig.

Lastly, here is a picture of the back of the brake.


Thanks,
Matt

uggh...the internets just killed my reply...

anyway, here's the gist of what I was thinking...

Thanks for joining the discussion, Matt! I think it's awesome you are putting this design out there.

What are your thoughts on...:

1) ...your front brake aero performance, at beta (or otherwise) , considering that you have chosen to "hog out" mass on the back side of the brake? The backside geometry you've chosen seems a bit draggy to me (at beta or otherwise).

2) ...how your brake impacts the fork's ability to create suction at beta?

3) ...your rear brake aero performance, considering that the "back" of the rear brake is essentially the leading edge from a flow perspective.

3) Also, your flow visualization modeling seems interesting. What CFD package are you using?

4) CAD limitations -> yeah, CAD packages of today can do some pretty amazing things from a surfacing perspective - here's what the guys I work with come up with:

http://www.taylormadegolf.com/#300S0

...just so I understand your thoughts above correctly, you chose to go with the stealth fighter look from a machine time/cost perspective at the potential expense of aero performance?

5) What is your perspective on the plots I've briefly thrown together above? yeah, I hear you...field testing is difficult business, eh? Hey, looks like the link to the docs/data here that is hosted in france:

http://www.simkinsdesigns.com/egg_drag_testing.htm

means that you've gotten the originator of "the chung method" (hey, when/how did it get that name, anyway??) himself, Dr. Chung, involved. What are his thoughts on the data he his hosting for you?

Anyway, thanks for stopping by, and Happy Thanksgiving, everyone!

kraig wrote:


This is where I'd like to see more photos with the rear brake mounted. If the profile is small enough, the back would be shielded by the frame?

Kraig,
1. The backside of the brake is tricky. I do not have much data to back anything up right now so hand waving will have to suffice. The most efficient design would probably be a more solid object. That is, if the material extended all the way back to the fork and frame, form the rear perspective, I think that is the best that you can do. A separate criterion was weight. I realize that the folks interested in this brake are not so concerned with weight, but I wanted a sub-300g brake to be on par with Dura Ace. Much to my chagrin, the brake ended up being just over 300g so something ended up being missed for the density calc in the CAD model. Anyhow, (here comes the hand waving) the rear brake should have less drag do to the small separation boundary at the arm/frame interface and the small cable barrel adjuster mount.

2. I will try to cover this in the computer-modeling article. The dynamics near the bike brake are especially complicated. You have allotted of geometry coupled with the surface of the tire, which is moving at twice the speed of the bicycle. With the exception of the drive train, the brakes are probably the most difficult area to understand flow.

3. The shape of the rear of the vehicle is often more important than the front. To be honest, the surface geometry at the front face of the brake was designed for frontal flow, not trailing. A more optimal rear shape would be more cone like, but the brake was also intended to be mounted under the chain stays and I wanted the overall standoff to be as small as possible.

4. Interesting work, too bad it is so difficult to machine that kind of stuff into the surface of a brake.

Regarding the data and the conditions between each run, the problem with that test was that the brakes had to be changed out between runs. This is the reason why the testing did not include the rear brake too; the change out time was too long. I liked you presentation of the data, Mat Lab right? It looked allot more concise than the excel sheets used in the report.

COSMOSFlow works express was used to model the brake. Unfortunately, I do not have the software myself (it costs a fortune) so that makes it a slow process to iterate the design. It is way better than trying to build dozens of prototypes.

Thanks for your thoughts, Matt.

Best of luck to you, and stop on by to let us know when you get your next write up done.

kraig wrote:


Since Matt kind of punted on that one...and for those following at home, here's some of my additional thoughts/concerns/randomness/provisional summarization ) regarding the files linked above based (partially) on new information acquired this morning in my garage (hey, new information is allowed, isn't it?):

[forgive the formatting ! ]

1) my "garage wind tunnel" test this morning was pretty interesting eh? - http://forum.biketechreview.com/viewtop ... f=1&t=2633

2) my "eyeball wind tunnel" is a bit confused (what's up, Jack Watts! ) ) at the magnitude claimed.


3) making assumptions and then fitting data to those assumptions just doesn't seem quite "right" to me - my personal, CAVEMAN, preference is to let these data speak for themselves, eh? (check out the upper right panel in my Matlab plot above...oh, that's an expensive plot...have you seen the renewal fees on a Matlab license!...slope is kinda like CxA and y-intercept is kinda like Crr. FWIW)

a) distance covered (for the same # of laps) for the files above is several meters different according to how I reduced the data - how much does this matter (in terms of delta CxA) for all the different methods of reducing "chung method" field test data that exist out there (there's lots of undocumented data reduction methods out there, eh) ?

b ) when implementing the WLB method, these data don't seem to support the assumptions made in the paper that Matt wrote and that Dr. Chung (Berkeley, is it?) is hosting.

c) the range of elevations that I observe between "good" files (while letting the data speak for itself without arbitrary manipulation/assumptions) concerns me - hence, my comment about conditions - perhaps the "steering angle" condition (which might affect Crr?) wasn't controlled? tough to say, eh?

4) I'm not sure how to handle the weirdness/discontinuities in the slope panel of my plot (data drops/repeats?!?) - y'know, 'cuz I like to let the data speak for itself and all...I will say, though, that my experience with coasting on these types of "picking the fly sh!^ out of the pepper" deals, is that you shouldn't intermittently coast when taking data with a power meter (SRM/wired PT based opinion).

5) multiple, independent, lines of inquiry is a good thing! )

[/forgive the formatting !]

What's everyone else think about this whole deal?

One thing that I think should be considered is that different bike companies design bikes differently. I think that Matt was trying to get the brake to fit "in front/in line" with the fork as much as possible. Cervelo has always used a smaller crown area, and so brakes probably have more of a tendency to stick out. However, Felt and I believe Trek use a little bit wider fork crown area, such that my stock brakes on my B2 fit "in line" with the fork except the two arms. I believe Matt brought the arms "in" but as we saw with the Cobb mod, at yaw the shortened brake arms didn't seem to matter.

I really think it's cool that Matt is trying to clean up the front of the bike. I have an inkling that the most gains would be had from designing to a specific fork, which is a less effective business model.

would be interested in seeing photo of rear brake mounted pref on a p3c ) ....but on any frame would be nice ...cheers