www.recumbents.com - Shell Design Software
www.recumbents.com
www.recumbents.com
Home | Profile | Register | Active Topics | Members | Search | FAQ
Username:
Password:
Save Password
 All Forums
 www.recumbents.com
 Technical / Bike Building
 Shell Design Software
 New Topic  Reply to Topic
 Printer Friendly
Previous Page | Next Page
Author Previous Topic Topic Next Topic
Page: of 23

shooky56
recumbent guru

USA
545 Posts

Posted - 12/29/2014 :  22:14:14  Show Profile  Reply with Quote
Reread what I wrote, phrased something poorly in regards to reducing frontal area. What I meant to say was that, when dealing with aero designs and comparing the benefits of one over another, the smallest frontal area is the best. The slight drag differences between them are so small that a fatter slicker shape with a better Cd ends up producing more total drag when multiplied by the frontal area. Tinier is better. At least that's what I came up with on paper.

The knee pocket designs of particular interest. Not sure how they play out in the racing but that would seem to follow that small frontal area paradigm by allowing the head to be lower and still be able to see.
Go to Top of Page

Speedbiker
human power expert

USA
3755 Posts

Posted - 12/29/2014 :  23:16:40  Show Profile  Reply with Quote
If that is so, why the liner that holds the world speed record is one of the larger designs? Actually, it's extensively laminar flow does in fact make up for it's greater frontal area. Now that I think of it, the 80 mph Cygnus also had greater frontal area than others. Just because you use the word "laminar", that doesn't mean you understand it at the level of current designers. Have you studied any aerodynamic design information that relates specifically to vehicles operating within our specific Reynolds numbers than has been written in the last 5 years? 10? Are your design parameters current to our specific aeeodynamics?
Go to Top of Page

shooky56
recumbent guru

USA
545 Posts

Posted - 12/30/2014 :  05:44:47  Show Profile  Reply with Quote
Larger designs: Do you know those numbers? I haven't been able to pull the cross sectional area info for any of these. But both the ones you mention look like incredibly cross-sectional areas. Again the "frontal area" as I use it, refers to the hole you punch with the entire shape (largest cross sectional area). In particular the Velox3 looks very small even using special power system to reduce the pedal box volume? (looks that way anyway).

Laminar flow: We may not be speaking the same language and perhaps I am and have been speaking different than everybody else in the forum. There is a lot of skewing of terms whereby one term is swapped for another, this is mentioned in the theory reading. For example, aspect ratio is often used for thickness ratio (a misstatement I also use). Everybody wants the laminar flow to be maintained as long as possible but a laminar airfoil curve is a particular curvature similar to the standard airfoil designed to place the point of maximum thickness further aft. A smooth shape designed to hold the turbulence as small as possible might also be called laminar also that isn't the way I have been using the term. Because my rider box is far aft of the entire shape and because the tail is "blunt" (this is for MY recumbent only, the software doesn't require this) I do, technically, maintain laminar flow far beyond the 1/3 total length (as does every fast HPV out there). So, technically this software will build a laminar flow vehicle if you put the fattest part further back than 1/3 its entire length.

Understand: I make no claims to understand any of it only have my opinion (and have stated such). My opinion is driven purely by numbers and will cameleon to any the numbers prove out. If you have a formula for the curvature of the Cygnus or Velox3 I'd be happy to incorporate it.

Vehicles within our Reynolds numbers: No sir, none at all. Without redoing my calculations, and if memory serves we run around a million? At any rate it seems like it was 500K to 1.5M, something in there. Very low Reynolds numbers make much of this discussion mute, of course and we (HPVs) do exist at a rather interesting transitional area. I suppose you point to it because of that?

Warren: As usual Warren has almost "sage advice". The entire drag profile of these is incredibly complex, far beyond my ability to calculate or even know HOW to calculate. Interaction with the road and, in particular, efforts to control the energy in trailing vortices, leads to a lot of aft curvatures that are particular to each individual shell. I am certain the racing efforts of the many teams and individuals will the racing shell into a very good form but I must agree with Warren that the best design is very individualized according to needs. My goal was to build a fast trainer, possibly something to do cross-country trips in, etc.



Go to Top of Page

Speedbiker
human power expert

USA
3755 Posts

Posted - 12/30/2014 :  09:07:46  Show Profile  Reply with Quote
As I suspected. Keep up the good work. Your tool will be useful to many builders.
Go to Top of Page

shooky56
recumbent guru

USA
545 Posts

Posted - 12/30/2014 :  10:38:35  Show Profile  Reply with Quote
You know Speedy, you have sent me looking (again, I did before I started this) for info about other HPVs out there. Their particular profiles appear (as best I can find) to either be unavailable. Can only assume this is because either the builders are protecting it or, as I suspect most, they just don't have that info. In fact, it appears Georgi (sp?) designs mostly by "that shape seems to flow" more than anything else. That is what I'd like to talk about, just how valuable a skilled eye is (or at least was 15 years ago). At the time, I was not on this project and couldn't discuss the details legally but can discuss the project in general terms as that is no secret. What struck me most was that, despite having multiple super computers at the design level, the shape was tweaked by a skilled aerodynamicist observing the pattern in a simulated wind tunnel and moving the surfaces accordingly. It took the marriage of the software and a human eye to produce the best results. Mostly these tweaks were at the transition zones where one shape joined another.
Go to Top of Page

Speedbiker
human power expert

USA
3755 Posts

Posted - 12/30/2014 :  14:05:19  Show Profile  Reply with Quote
Every liner that has neared 80 mph, and there are several, are priority, math driven shapes. Some, such as ours, were treaked with pro level cfd. Others, like Delft, were tweaked with cfd, wind tunnel, and track testing. Every builder at this level has the numbers you seek. You are correct that few of these guys want to share info. Although sometimes it has to do with contractual obligations, usually it is because they frequently get asked for such information by people who are self taught wannabe aerodynamic designer who will not begin to understand the info given. And what follows is usually endless questions whose answers are beyond their grasp. Worse yet is that the people asking for info often want to tell these guys who have aerodynamic engineering degrees, years of experience, and several record setting builds under their belts, how to improve their designs! If it sounds like you might have to qualify to ask these guys for cutting edge data they worked hard and long for, you are right. What formal studies or degrees do you have in aerodynamics, or what competitive vehicles have you designed and built? This isn't meant to dismiss you, but rather to let you understand that guys who have spents years in school and the field do not share their knowledge with just anyone. But, of course they have the data!!
Go to Top of Page

alevand
human power expert

USA
2931 Posts

Posted - 12/30/2014 :  15:00:16  Show Profile  Visit alevand's Homepage  Reply with Quote
The Varnas were pure sculpture, no cfd involved.

C:
Tony Levand
Go to Top of Page

shooky56
recumbent guru

USA
545 Posts

Posted - 12/30/2014 :  15:20:01  Show Profile  Reply with Quote
Thanks Tony. Tony I have read your build Leviliner(sp?) build thing and marvel at the parallel thought paths we have taken. You are exactly modeling what this software does when you have taken the points and extrapolated them. Even using the same formula series. This doesn't do it that way (using the formulas directly) but the results are the same. I'm not using S-Splines but they'd be easier than the Beziers used for the tug shapes. Wish I'd thought of that, no reason to replace something that works though.

Speed: I'm surprised those guys would have reluctance to discuss their projects. I've never known a good engineer that wasn't bursting at the seams to discuss their project.
Go to Top of Page

Speedbiker
human power expert

USA
3755 Posts

Posted - 12/30/2014 :  15:46:53  Show Profile  Reply with Quote
The Varna is the famous, genius shaped exception. Of course Georgi doesn't have CAD data! Is that who you asked for data from? Have you even attempted to contact David Weilemaker, Thomas van Schaik, Damjan Zabovnik, Charlie Ollinger, or any of the top designers? Who did you ask? Anyone? Or are you complaining about fictional persons not willing to share data? I suppose you expected guys building state of the art competition machines to publicly post their designs? Perhaps you thought if you posted your project online that the world's top designers would jump right in and tell you what they know. Maybe Tony will FTP his 3D CAD file and CFD files to you.
Go to Top of Page

warren
human power expert

USA
6117 Posts

Posted - 12/30/2014 :  17:24:04  Show Profile  Visit warren's Homepage  Reply with Quote
Thom - The software Wayne is designing is not intended to compete with the high end, big $$ Battle Mountain teams, it is designed to provide a tool to allow the creation of a fast, custom fit fairing with a lot of artistic freedom.

Designing a fairing is a lot like religion. Everybody has a different idea of what the one true religion is. Some religions are more popular than others. I think Wayne is designing a non-denominational religion fairing design tool. This is a good thing.

Keep on coding Wayne!

-Warren.
Go to Top of Page

Speedbiker
human power expert

USA
3755 Posts

Posted - 12/30/2014 :  17:36:03  Show Profile  Reply with Quote
Right on. Like I said above, he is creating a tool that will be useful to builders.
Go to Top of Page

alevand
human power expert

USA
2931 Posts

Posted - 12/30/2014 :  17:45:38  Show Profile  Visit alevand's Homepage  Reply with Quote
I don't know why you are jumping up and down, Thom. I think what Wayne is doing is cool. Need a way to make the form. I know that you can get custom architectural foam shapes for stucco, but I don't know if they are true 3d. I think you could get a rough hot wire cut shape by using 2d plus rotation. If it were a round cross section, then so may facets could be cut. This would work for an elliptical cross section also, if not too extreme. The rotation could be manually indexed.

https://www.youtube.com/watch?v=yKpq9FZMPqQ

C:
Tony Levand
Go to Top of Page

Speedbiker
human power expert

USA
3755 Posts

Posted - 12/30/2014 :  18:15:05  Show Profile  Reply with Quote
What is your shell going to be made from, Tony? Composite, please!!!
Go to Top of Page

Speedbiker
human power expert

USA
3755 Posts

Posted - 12/30/2014 :  18:23:09  Show Profile  Reply with Quote
What is your shell going to be made from, Tony? Composite, please!!!
Go to Top of Page

shooky56
recumbent guru

USA
545 Posts

Posted - 12/30/2014 :  18:58:32  Show Profile  Reply with Quote
Thom: No sir, I haven't attempted to contact anyone. Was just remarking about your comment that they don't like to talk to unqualified people. The discussion about the top designers was initiated by you not me, I don't care at all, have no intention to race or any of that. My remarks about their data was initiated by your statement of the Varna and Velox3 having among the largest cross-sectional areas out there and, because of that I figured you had that data and was asking you for it, that is all. Was curious what kind of cross sectional areas those guys were ending up with.

I don't normally engage conversations that are aimed at the person rather than the topic and today wasn't particularly productive. Probably need to quit side tracking myself and keep plowing forward. Got to get that crank reconnect to the foot!

If you guys think you'd like to play with the frame builder I can prepare another MSI. The frames you make are savable and should be loadable with later version unless I severely change the format (quite unlikely at this point).
Go to Top of Page

Speedbiker
human power expert

USA
3755 Posts

Posted - 12/30/2014 :  21:54:29  Show Profile  Reply with Quote
I didn't say Varna. I said Velox and Cygnus. None the less, keep working on that program.
Go to Top of Page

shooky56
recumbent guru

USA
545 Posts

Posted - 12/30/2014 :  22:37:19  Show Profile  Reply with Quote
Ah yes that's right and I will, thanks Thom.
Go to Top of Page

jjackstone
recumbent enthusiast

USA
232 Posts

Posted - 12/30/2014 :  23:17:54  Show Profile  Reply with Quote
A tool that can help determine some of the liner's frontal area can be found at the Human Power The Film website. Photos of a number of the bikes raced at BM have been taken with a numbered grid behind the bike. While you won't be able get a precise area, a good estimate can be made from these photos of a number of the bikes. Here is a link: http://humanpowerthefilm.com/gallery2012/human_power_the_film/index.html
It should take you directly to the page I described. If not, go to humanpowerthefilm websight. Then click on gallery. Two subgalleries show up. Click on the gallery2012 choice. Hope this helps. Some bikes you may recognize, others you might want to cross reference with the WHPSC photos found on this site.

JJ

Edited by - jjackstone on 12/30/2014 23:20:00
Go to Top of Page

alevand
human power expert

USA
2931 Posts

Posted - 12/31/2014 :  05:58:14  Show Profile  Visit alevand's Homepage  Reply with Quote
quote:
A tool that can help determine some of the liner's frontal area can be found


Cool.

The Varnas are really beautiful.

My output file was a set of point coordinates that I converted to an IGES format, the machine shop could read into their Cam program. I may have given them splines, I don't remember now it was 6 years ago. I had a quote, something like $8k to cut the foam.

C:
Tony Levand


quote:
What is your shell going to be made from, Tony? Composite, please!!!


No, something much better, I cant tell you now, but you'll see.

Edited by - alevand on 12/31/2014 06:53:02
Go to Top of Page

shooky56
recumbent guru

USA
545 Posts

Posted - 12/31/2014 :  07:02:09  Show Profile  Reply with Quote
Thanks JJ! Actually I can analyze it pixel by pixel if I have the scaled frontal shots and had thought about doing that anyway! Way cool sir.

Foam cutter: I started to build a 12' by 8' foam cutter a few years ago and got side tracked and never finished. Not saying I'll finish it now but I was only working on about 1/4" precision anyway 3-axis plus a rotation "spicket" the slow speed (3600rpm) router shaft is 2 feet long so it wouldn't really be a 5 axis but could still cut the ends properly.

I'd never charge anything like that, in fact for the community I'd do it for the nuisance and mail handling but that's probably mute, at least for the next few years.

Go to Top of Page

Speedbiker
human power expert

USA
3755 Posts

Posted - 12/31/2014 :  07:10:47  Show Profile  Reply with Quote
Finish up that 12 foot foam cutter! It's every hpv builder's dream. And stepper motors and controls are dirt cheap now days.

Whoever thought up that liner scaling deal was a genius! Ha ha ha...
Go to Top of Page

Upright Mike
human power expert

USA
3803 Posts

Posted - 12/31/2014 :  07:38:38  Show Profile  Reply with Quote
quote:
Originally posted by Speedbiker

Finish up that 12 foot foam cutter! It's every hpv builder's dream. And stepper motors and controls are dirt cheap now days.

Whoever thought up that liner scaling deal was a genius! Ha ha ha...



Try looking at a TechShop location near you. They allow access to a Shopbot 4 foot by 8 foot CNC cutters that can cut foam and wood. I was a member at the Detroit facility. Thom and I toured it, and our Michigan human powered vehicle meeting was held there last year. You can import CNC code, and they have lots of classes and equipment for DIY Do It Yourself makers, plus a staff of Tech aides to help you.
http://www.techshop.ws/locations.html

Based off of Thom's idea which Warren posted on this streamliner scale page.. http://www.recumbents.com/wisil/scale_project/streamliners.htm. I created the Battle Mountain One-Inch Scale Board. A duplicate board is in the works to be used at midwest events. The side views give reasonable accuracy, of the vehicle's side shape, when the camera is about 40 feet away, based on my own experiments with my own streamliner. A zoom lens is used. This helps reduce the parallex error so the shape is not growing in length when projected onto the grid board. In SIDE views the fattest part of the vehicle shape is only about about two feet or less away from the board. However, I WOULD NOT TRUST THE FRONTAL VIEWS. In FRONTAL views, the shapes are "growing" because their fattest part is much further, like four to five feet from the board, so they look bigger. I had plans for another board with a big cutout that could be used for frontal views and placed at the biggest frontal area, along the vehicle length, but never finished that. THis would be more accurrate. If I had time and money I would buy a dozens of laser pen lights, build a stand to make them all one-inch apart and have them shine perpendicular onto the surface at the widest point.

I made that board in 2011, hurt my back (but not as bad as you Thom in 2011!) from leaning over it for hours drawing the lines. It folds up into 2 foot by 4 foot tall panels. Making all those panels lineup was tricky. The six inch line near the bottom, you can see I drifted. I shipped it out there at a high cost, since i couldn't attend. I talked people into using it, and the Human Power crew took it from there. In 2012 it was used again, but not since. I believe it's kept in storage at the Civic Center, not in the other storage facility outside town or in Reno with other WHPSC gear.

Edited by - Upright Mike on 12/31/2014 08:14:31
Go to Top of Page

Upright Mike
human power expert

USA
3803 Posts

Posted - 12/31/2014 :  08:20:06  Show Profile  Reply with Quote
I have seen some really cool stuff used at work to scan surface vehicle shapes (including equipment that projects a grid onto the shape then measures the shape from the distortion of the lines or shadows). Another system paints the shape with hundreds of passes with a laser. To do a streamliner shape, you really need a scanning 3D system, that will automatically capture thosands of points at a somewhat regular interval that can be put into a CAD program and splines grafted between the points.

Many places have 3D measurement arms like a Romer or Faro arm or Trace Cam wand (all of which I use regularly at work). These are good for maybe ten points or less. But it would take a hundred hours and you would drive yourself crazy keeping track of where you are or where you been on your streamliner surface while tracing it, unless you drew up a grid yourself on the streamliner shape. You would end up with points not at a regular grid interval, and I think the resulting shape would be hard to spline fit. Now only if I could have a lab to myself over the weekend, and sneak my streamliner in. Of course we got some wind tunnels too.... but damn I'm just a lowly lab rat with not enough pull to ask for such toys. Though I did get our entire repaved 5-mile high speed track for a three-day human powered record setting event in 2009! :-)

Edited by - Upright Mike on 12/31/2014 08:23:24
Go to Top of Page

shooky56
recumbent guru

USA
545 Posts

Posted - 12/31/2014 :  10:08:16  Show Profile  Reply with Quote
Thom: To make matters worse I BOUGHT everything to build it. The biggest steppers (forget now and would take me a bit to find the vendor, the Gecko controller people though I think had the steppers too). Steppers are KL34H2120-42-8B's Drives are 6203v (no longer in production), believe those were replaced by G540s
Drives:

http://www.geckodrive.com/geckodrive-step-motor-drives/g540.html

Motors:
http://www.automationtechnologiesinc.com/products-page/stepper-motors/nema-34-high-torque-stepper-motor-1200-oz-in-half-inc-single-shaft-with-flat-2

The darn toroidal transformer power supply probably puts my life at risk with all the copper thieves these days! But it is a heavy chunk of transformer (just guessing 20 pounds or so).

Mike: Just thinking about a method to marry a point dump of your scans. I really do think I could S-Spline the rest of the points in between, it'd be a 2-D spline effort and would take some head scratching to get right but it seems math-doable. May decide differently 2 days in and even fewer hairs on my head. Very interesting stuff though.

JJ: Boy oh boy you are EVIL!!! (lol jk, of course). That link has sooo sidetracked me. Whoever did all that work put a lot into those shots, it's too bad they couldn't shoot across a gym floor longways to minimize the parallax. Thought about estimating the camera distance but don't know that I really care too much about that, most of the fattest parts are further back and, for most of the small models, the fat part is similarly positioned and so they all have the same error. Something that impressed me about the Varna was it was just about the least laminar design out there (its fat part being much forward of the typical) and therefore should have a more effective tailbox. Of course we'd be able to extract all the Cd numbers if we knew the motor size, something that varies from rider to rider of course.



Go to Top of Page

shooky56
recumbent guru

USA
545 Posts

Posted - 12/31/2014 :  13:39:00  Show Profile  Reply with Quote
Ok analysis done. Did some estimations based on a little bit of known data, of at least I think known. The exact square inches will suffer some from my parallax corrections which were calculated thusly.

The Varna appears to be 19 inches wide, and (if my info is correct) is actually on 16 inches wide. The error is ~20% but I used that exact ratio. I do not have enough solid info to even give a rough estimate of my possible sizing error but I suspect it is less than 20% (for absolute square inches of area).

The vertical parallax is a bit worse because the lowest part of the nose was much closer to the camera than the fat part of the width. I finally gave in and used the same for the vertical parallax which means, if anything, I am overstating the total area. Worst of all each model probably has a slightly different error, meaning that ratio comparisons aren't precise either.

With all those caveats, the models I could identify and analyzed were the black Varna (picture 19?), the Velox (Velox 1 I think but not sure), and the Wedge.

The Wedge looked really tiny but fell behind the Varna and, as Thom said, the Velox had a huge (relatively speaking) cross-section.

in units of square inches "i2"

Varna: 113 i2
Velox (1?): 158 i2
Wedge: 127 i2

Also, I do have several comments from perusing pictures since then (mostly looking for frontal or top down shots of the Cynus3). A picture showing what I think is a Velox3 half-shell lying on the ground looks like the thickest part is quite a bit forward, meaning it isn't laminar but more "standard airfoil". The Velox in the gallery was definitely laminar.

The Varnar also surprised me that the thickest part appears to be about 40% chord length behind the nose. 25% is standard airfoil, 50% or more is laminar, the Varna exists between the two.

Also a correction: I'd said 33% length was standard airfoil, something I'd remembered crooked. 3 parts are behind the fattest part and 1 part in front for 1:3 ratio (which means fattest is at 25%).

...and one last thing. There's no hard theory here just an impression based on many unrelated little quirky trivia I've accumulated over the years. I remember that "anything sticking out was hideous" for drag. Canopies are surely terrible but what are you going to do, I'm not building a trainer without a bubble as I know of three people that have been killed by motorists in my area alone (2 were recumbent trike riders). But for racing, I can see that anything goes. The fat cross section of the canopy-less Velox line most likely means that "it was worth it to remove the canopy".

BTW one of the quirky things was a recent well-manned, well funded effort in California to produce a 3-wheeled EV.

http://en.wikipedia.org/wiki/Aptera_Motors

On their website they touted that "our total drag at 60? mph is less than that of a truck mirror". There's a lot of questions in there as there are a lot of truck mirror types but the point is clear, things that hang out are bad.
Go to Top of Page
Page: of 23 Previous Topic Topic Next Topic  
Previous Page | Next Page
 New Topic  Reply to Topic
 Printer Friendly
Jump To:
www.recumbents.com © 2017 www.recumbents.com Go To Top Of Page
Snitz Forums 2000