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 - 03/10/2015 :  17:40:58  Show Profile  Reply with Quote
lol yup Jeff. That high rider sucks for a shell though.

Couple things, I close my tailbox to an edge, which isn't bad per se but intuitively I fell (but do not know) that there really should be a bit of reverse curvature, kind of like after the bulge of a guppy it tapers but the curve goes backward to feather. I mean its "just the way stuff looks". But, at least, the tailbox is dirty so probably less valuable to be perfect. At any rate made good progress on the rib mathematical definitions. Still some work and I need to add special ribs at the front and tail that are in X and Y planes (these are all in the Z plane).

Also need to add the "spar holes" for pipes or 2x4's or "something" to hold these spaced and in alignment. Even thought about making plywood sections with cut-out slots sort of assemble it like cardboard models (insert tab A-A into slot A-A).

[URL=http://s1372.photobucket.com/user/shooky56/media/Cycling/Ribs%20still%20a%20bit%20of%20work%20to%20do_zpsw02jg6aa.png.html]
[/URL]
Go to Top of Page

Speedbiker
human power expert

USA
3755 Posts

Posted - 03/10/2015 :  18:16:19  Show Profile  Reply with Quote
A recurve at the tail helps with pressure recover, but only if conditions are perfect enough to allow controled flow all the way back. The other 99.9% of the time the recurve sets up more turbulence at the tail.
Go to Top of Page

shooky56
recumbent guru

USA
545 Posts

Posted - 03/12/2015 :  05:19:13  Show Profile  Reply with Quote
Again (on the correct page this time), thanks for the recurve comment. Have taken that concern off my list of worries on this thing.

Never thought it'd be so hard to slice this thing up. It really wasn't hard, just hard to get right. At any rate, want to add a couple of nose pieces that lie in the XY and XZ planes and then I'll be ready for output. Having criticized MicroSoft for their mixed bag of references, I now find myself counting on them to simplify the next step (after the couple of nose pieces). Some of these things are defined by as few as 60 points which will make the perimeter "unsmooth" (like using an octagon to define a circle). MicroSoft has a built in smoother to draw function. You hand it a set of points and it'll draw a smooth curve.

I'll need to build an interface for the setup which would include the rib spacing, scale (maybe you want to build a model rather than full-scale) and also what bit you'll be using on the CNC. Probably need a preferred spar type (maybe anything from 3/4" conduit to 2x4's

Have thought about the bit...<cough> a bit too. It'd be nice to have a bevel cut on the rib to match the shell surface but what I think I'm going to do instead is make the rib correct at the largest side. That means the rib would stick out where its thickness goes into a slightly smaller part of the shell. The crafter can use a belt sander (or something) to sand the high edge sloped toward the next or previous rib. I envision a long jig hooked to a belt sander that touches the fore and aft ribs to limit the sanding angle. I have the strongest suspicion that anybody doing one of these is very intimate with sanding techniques.

Anyway here's a top down view of the same shell with "good ribs"

[URL=http://s1372.photobucket.com/user/shooky56/media/Cycling/Ribs%20getting%20closer_zpsqnlo7dwf.png.html]
[/URL]
Go to Top of Page

warren
human power expert

USA
6117 Posts

Posted - 03/12/2015 :  05:53:52  Show Profile  Visit warren's Homepage  Reply with Quote
Your app is beginning to bear fruit. Cool images!

Is virtual Warren missing his right leg and arm intentionally?

Edited by - warren on 03/12/2015 05:56:49
Go to Top of Page

shooky56
recumbent guru

USA
545 Posts

Posted - 03/12/2015 :  07:35:20  Show Profile  Reply with Quote
Right leg: Yes he is Warren, much as I didn't bother to do the "back side of the shell, I didn't give you a right leg either. <snicker> Buck it up, take another aspirin and keep pedaling! Hopefully I'll get around to adding all of that, and it shouldn't be too difficult but it does add a bit of complexity that isn't necessary to finish the shell. The detail was the "other pedal" which is (of course) 180 out. Since all of the kinematics references a "global pedal location" I'd have some work to clean up the single pedal references and expand it for two. Easier would be to tie both legs to the same pedal (respecting their left right positions) so both knees were up at the same time, but hopefully I'll get around to fixing that back leg sometime.

I studied this a while, and think I'm going to stop here with the ribs and work on output. Leave for bro's in the AM. Dang it, so close to releasing a version too. Oh well, all in good time.

Here's a picture showing the latest nose rib added, also put it at an angle so you can see the gap which would be filled by 1/2" plywood in this example. Did that so I could make sure the shape accounted for material thickness.

[URL=http://s1372.photobucket.com/user/shooky56/media/Cycling/Ribs%20before%20output_zpstjgniodc.png.html]
[/URL]
Go to Top of Page

shooky56
recumbent guru

USA
545 Posts

Posted - 03/12/2015 :  09:52:52  Show Profile  Reply with Quote
Starting on the UI for the Output. Decided to add a pop-up. The real estate was getting crowded. Just captured the left edge of the popup which has all the input area on it. The large 2" rib thickness was just to make sure the nose rib was honoring it properly. Appears to be correct.

[URL=http://s1372.photobucket.com/user/shooky56/media/Cycling/Rib%20UI%20Window%20-%201_zpsa9effu0f.png.html]
[/URL]
Go to Top of Page

shooky56
recumbent guru

USA
545 Posts

Posted - 03/22/2015 :  15:26:46  Show Profile  Reply with Quote
Quick update. The family socializing has put a big damper on progress but I'm still on it. Ready to work on sending it to a printer and adding alignment holes.

[URL=http://s1372.photobucket.com/user/shooky56/media/Cycling/Ribs%20ready%20to%20add%20print%20function_zpsiclgpl5b.png.html]
[/URL]
Go to Top of Page

shooky56
recumbent guru

USA
545 Posts

Posted - 03/23/2015 :  18:14:13  Show Profile  Reply with Quote
Nephew went back to San Antonio so that gives me more time in mid day even though the nieces seem to think a minute of solitude is a sin... "bless their hearts". At any rate I'm getting a couple hours in a day coding. Had some happen I did not expect. While printing life sized ribs for the printer dump I noticed a "dip in the hood" on one. The dip may remind you of the shells with knee humps although not nearly so pronounced and I was scratching my head as to why. Then it dawned on me. Even though the rider box is enforced to no reverse curve (i.e. a heart shaped cross section). What happens is, with a high lean angle to the rider box and squarish pulled shoulder points the extra length to the Adam's apple from the nose makes a gentler aero profile and actually dips below the curves headed to the shoulder. Surprised me when I saw it but it looks like it's doing the right thing.

The reason I worried about interior curves isn't strictly all that important, it was so that a split mold would pop if the splits were side halves as opposed to top and bottom. So I was thinking about fixing it and the old adage "if it ain't broke don't fix it" popped into mind. Really it isn't broke you should be allowed to violate that curve with the understanding that you're going to split this thing top and bottom. Eventually I may give a switch for letting the software confine the shape for it but, in the interim, I'm going to allow the designer to do what they like. If their ideal shape has those kind of curves, something as thin as a shell would probably pop anyway. They could always simply sand the curve out too although the airfoil integrity would suffer a bit.

So now I'll move on to adding alignment holes and CNC codes. Brother said the CNC hasn't been fired up for a while but worked last time.
Go to Top of Page

warren
human power expert

USA
6117 Posts

Posted - 03/24/2015 :  05:49:17  Show Profile  Visit warren's Homepage  Reply with Quote
Hi Wayne,

Do you know yet what bike you will be putting in the body you are making for your yourself is, and what the resulting body looks like?

-Warren.

Edited by - warren on 03/24/2015 05:49:40
Go to Top of Page

Speedbiker
human power expert

USA
3755 Posts

Posted - 03/24/2015 :  07:36:13  Show Profile  Reply with Quote
While this is a fascinating design exercise I hope the design you posted is not the vehicle you intend to build. While I understand you just want something of a practicle enclosed vehicle, those of us who have designed and built many liners can see that the design will yield poor performance. And after hundreds or even thoudands of hours of design and construction, an under performing vehicle is heartbreaking. We've seen it many times on this site.
Go to Top of Page

warren
human power expert

USA
6117 Posts

Posted - 03/24/2015 :  08:56:39  Show Profile  Visit warren's Homepage  Reply with Quote
No Thom, that is a test shell based on me on the high racer I built for my brother.
Go to Top of Page

shooky56
recumbent guru

USA
545 Posts

Posted - 03/24/2015 :  09:30:13  Show Profile  Reply with Quote
Yes Thom, Warren is right and something to the effect of "not a practical design" is mentioned several times (not that I expect anyone to wade through all of that). In fact I intentionally built "something I'd never build" to make sure I was honoring designs outside my original concept. It's easy to write software with blinders, designing a shell that is unusual doesn't mean the software will design all shell types but it helps keep my thinking out of the gutter.

My worst fears is that my software will build a "shell that doesn't fit" and it is to that end I've put most of my effort. In fact, for my own, I intend to do cardboard ribs, cut their middles out so they look more like the bracing inside an aircraft fuselage and make sure I fit properly before cutting the plywood. Unfortunately I have to mock up a frame out of wood because I left mine back home.

Also it's important to note that I never even considered a world class shell, didn't before and am not now. Only a few of the build projects in here are intended to be "the fastest" most are just somebody enclosing a bike or recumbent. The intent was at the start and remains "hey this ought to be a lot better than guessing" for an aero shape.

Now, that said, it's possible that it can be used for very fast designs. In fact I expect that it would make fast shells if built around small footprint frames.

As I near a rear run on the CNC I'll have to dig a little bit to find my measurements for my frame. They are built in to versions way back there somewhere.

Go to Top of Page

shooky56
recumbent guru

USA
545 Posts

Posted - 03/24/2015 :  09:37:11  Show Profile  Reply with Quote
Oh Warren, yes I do know the bike. My real frame was built in to the first release although the software wasn't capable of rendering a trike at the time.

Built my trike on the frame builder but the save formate changed so drastically I'll have to do it again, shouldn't take too long but I'll be pretty careful about accuracy because it's going toward "a real shell".
Go to Top of Page

Speedbiker
human power expert

USA
3755 Posts

Posted - 03/24/2015 :  10:09:31  Show Profile  Reply with Quote
Snook, I know that you aren't trying for a super fast design. I just don't want to see you sad if all your work produces a vehicle that is no faster and much less managable than before adding a shell. I hate to tell you how many times we've seen it.
Go to Top of Page

warren
human power expert

USA
6117 Posts

Posted - 03/24/2015 :  10:18:29  Show Profile  Visit warren's Homepage  Reply with Quote
Your frame builder will support trikes too?
Go to Top of Page

Speedy
recumbent guru

USA
883 Posts

Posted - 03/24/2015 :  11:28:41  Show Profile  Reply with Quote
Gotta say ...
your software looks very cool and hopefully it will get plenty of use



Steve Delaire

http://molten3d.blogspot.com
Go to Top of Page

Upright Mike
human power expert

USA
3803 Posts

Posted - 03/24/2015 :  18:08:51  Show Profile  Reply with Quote
Here is perhaps the most practical commercially available two-wheeled streamliner in the world today. The POB Peregrin on Birk streamliner from Birkenstock Bicycles. I think this makes a good model for someone to follow if they want a practical streamliner.
http://speedbikes.ch/peregrin-on-a-birk/

It looks like a fantastic machine with bomb-bay foot doors, tilt up shell for easy entry, head pod for excellent visibility. You can also get extras like rear view mirrors, even lights and turn signals. You can purchase the bicycle inside, a Birk Comet separately. It is a great carbon fiber recumbent with under the seat suspension.
Go to Top of Page

Upright Mike
human power expert

USA
3803 Posts

Posted - 03/24/2015 :  18:26:13  Show Profile  Reply with Quote
The Peregrin on Birk (POB) is also a fast machine. There are only a few of them in the world from what I can find (1 in Brisbane Australia for sure!), so not many speed marks from competitions that I can find yet. And of course some owners, maybe won't ever race them.

This past summer Charles Henry, rode 45.894 miles, 73.859 km in One hour at Dekra becoming the 22nd fastest in the world for the One Hour event. His speed was about 5 to 7 mph faster than any three-wheeled velomobile mark by other other strong European riders. I have yet to figure out if anyone has done a serious Top Speed run over 200 meters in competition with a Peregrine on Birk. But I know that earlier Birkenstock machines have gone over 50 mph (80 kph) on several occasions at European events.

On the much faster and longer Battle Mountain course, I think the Peregrin on Birk could easily be a 65 mph maybe 70 mph streamliner. {I base this on the fact that Gareth Hanks has gone nearly the same 45.8 mph in one hour for the Trike world record in the Completely Overzealous and he's gone 72 mph at Battle Mountain because he's got a lot of sprinting power)
Go to Top of Page

Upright Mike
human power expert

USA
3803 Posts

Posted - 03/24/2015 :  18:28:55  Show Profile  Reply with Quote
Here is member Stix's excellent writeup on his POB Peregrin on Birk
http://www.recumbents.com/forums/topic.asp?TOPIC_ID=5251&SearchTerms=POB
Go to Top of Page

Joel DIckman
recumbent enthusiast

USA
129 Posts

Posted - 03/25/2015 :  06:27:52  Show Profile  Visit Joel DIckman's Homepage  Reply with Quote
quote:
Originally posted by Upright Mike

Here is perhaps the most practical commercially available two-wheeled streamliner in the world today. The POB Peregrin on Birk streamliner from Birkenstock Bicycles. I think this makes a good model for someone to follow if they want a practical streamliner.
http://speedbikes.ch/peregrin-on-a-birk/

It looks like a fantastic machine with bomb-bay foot doors, tilt up shell for easy entry, head pod for excellent visibility. You can also get extras like rear view mirrors, even lights and turn signals. You can purchase the bicycle inside, a Birk Comet separately. It is a great carbon fiber recumbent with under the seat suspension.



Hello Mike,

Stix (Glen) of Brisbane has had the Peregrine on Birk bike for almost two years now, and has also been riding a Lightning F40 fairing on a Lightning R84 bike for a longer time. I am not sure that he would call the Peregrine street streamliner the more practical bike though. Would like to hear from him about this. When he last compared them he said the Lightning was better for starting, turning, feeling less confined, and easier to handle in gusting crosswinds. Although he did not mention it, I am sure that the Lightning F40 costs a small fraction of the cost of the Peregrine on Birk.

(Some of us do not have deep pockets.)

He said the Peregrine was faster, at least in the right conditions. And it is certainly prettier, by a considerable margin.

But more practical? Maybe Stix will revisit this question for us, with more miles under his belt since his last comparison.

Safe riding, and green with envy,
Joel Dickman
http://lightningriders.com

These three prevent most accidents: seeing, being seen, & (usually) common sense.



Edited by - Joel DIckman on 03/25/2015 06:32:51
Go to Top of Page

Speedbiker
human power expert

USA
3755 Posts

Posted - 03/25/2015 :  07:31:17  Show Profile  Reply with Quote
Correct on all points, Joel.
No way that thing breaks 65 at BM.
$$ per mph is off the chart.
Go to Top of Page

shooky56
recumbent guru

USA
545 Posts

Posted - 03/25/2015 :  09:59:18  Show Profile  Reply with Quote
Thanks all for the kind remarks.

Yes Warren it will build anything except multiple cranks. The body movement is tied to the crank, I could designate "crank 1" as the animation but for a tandem I'd really need to animate a second rider. Don't figure that's much in demand. Of course anything can be done given enough time.

Can't do rear facing designs either although one could conceivable just move the rider box far forward. Rear facing kinematics won't work right though, the body would have to be flipped over and there's no provision for that in the software.

Speedy, thanks. Adding the alignment holes this morning then it's on to CNC stuff. Actually want to build and shell an upright to make sure it works for Mike too. It should but those leg angles will "flip quadrants" (their angles move to a different Cartesian quadrant) and there are dozens of fixes for what quadrant these appear. Many transcendental functions are defined in only two of the four quadrants and I manually fix the problems that produces so an upright could show me several errors.

Rather interesting thing yesterday. I thought about doing the print from inside the software and opted for the easy way out. The software creates image (*.png) files and you scale those to be life-sized (or model sized) if you prefer. The one thing you'll need to know is your printer's DPI. I included Horiz and Vert DPI because bro's printer was 72DPI but apparently the rollers didn't quite roll fast enough and it came out more like 76DPI vertical. Funny thing was I was unable to get that stat off Canon's technical specs. It said 1400 DPI scanning, 200DPI fax, stuff like that but no printing DPI. But just print a 400x400 square on a page and measure it. 400/measurement(in inches) = DPI. [400 because 400x400 is big enough to minimize error and still small enough to fit on a page for most printers]

Go to Top of Page

stix
recumbent enthusiast

Australia
136 Posts

Posted - 03/25/2015 :  23:38:22  Show Profile  Reply with Quote
Due to the number of comments in this thread about the PoB I have updated the PoB thread do reflect all the mods I have made to the PoB. I have also answered some of the questions posted here. Here is the link to the PoB thread.

http://www.recumbents.com/forums/topic.asp?TOPIC_ID=5251&whichpage=2

Stix

Peregrin on Birk (Birk Comet RT with full fairing)
Lightning F-40 (built around R-84 frame)
Soma Smoothie Upright
Go to Top of Page

shooky56
recumbent guru

USA
545 Posts

Posted - 04/02/2015 :  07:12:32  Show Profile  Reply with Quote
Not many will find interest in this post but I'm putting it in here just to show how weird the "tiniest details" can get.

Had a small problem. Bro worked with me on it last night. The problem arose in that my cross-section defining "rider's box" uses Bezier curve smoothing AND the rider box leans backward so the top center airfoils are longer and thinner profiles than the forward points. What this creates is a fold in the forward ribs like shown in picture #1. Picture #1 is greatly exaggerated as the fold is really on the order of 1/16th of an inch.

The resulting half-mod has an interior angle in it (picture #2). The forums say "don't do that". It bothers me quite a bit so I set about to fix it. In the course I found a couple of point set creation errors. I was handing MicroSoft duplicate points which makes its smoothing algorithm behave erratically. Fixed all that then set about fixing the dip.

My first solution (fairly arbitrary) was to patch the slopes between the points as they closed on the center line. Given enough tweaking that kludge could probably work for my shell. As I flush that code it was still visually obvious where my fixes occur.

My second attempt seems better. Aircraft designers of old probably had lots of rules of thumb but one of them was "when in doubt use an elliptical shape". I needed to come up with a way to "ellipsize the top".

Pencil and paper time. Took me a while to get there but basically an ellipse of the form

x^2/a + y^2/b = 1

has a slope at x,y equal to b^2*x/a^2*y

b^2/a^2 is constant, to clean up the algebra, I'll call it "k".

A bit out from the center, the shape looks perfect. My idea is to fit an ellipse to the "good points" and use that formula to overwrite the bad points near the center. I can come up with (u,v) and (u',v') with simple code. I can also easily approximate the slope at both points by using the bounding points (shown in picture #3). Given those two points and their slopes and using "S" to designate slope:

S' = k u'/v' and
S = k u /v

The problem is I know the x values because the Y line is the same both for the shell and the ellipse were trying to derive. The location of the X line for the ellipse is unknown so v and v', although they have a value, aren't related to the formula we're trying to derive. Undeterred...

shifting stuff around a bit:

v' = (k/S') u'
v = (k/S ) u

Subtracting gives:

dv = du (k/S' - k/S)

"dv" saves the day. Although v and v' correct values are not known their difference is. That means we know everything in the last equation but "k". Sparing the algebra (unchecked):

k = (dv/du)(S*S')/(S-S')

That's all I need.

Headed to code.

[URL=http://s1372.photobucket.com/user/shooky56/media/Rib%20problem%201_zps7zbwcspd.png.html]
[/URL]
[URL=http://s1372.photobucket.com/user/shooky56/media/Rib%20problem%202_zpsru9wvkip.png.html]
[/URL]
[URL=http://s1372.photobucket.com/user/shooky56/media/Rib%20problem%204_zpsasptm7ej.png.html]
[/URL]
Go to Top of Page

shooky56
recumbent guru

USA
545 Posts

Posted - 04/02/2015 :  07:20:28  Show Profile  Reply with Quote
Oops meant to click preview, got the wrong button. At any rate, I'm still in GA, still coding, getting feverishly close. Also getting a bit more time now that I've been here 2 1/2 weeks. Still have the obligatory dinner and spades tourney at a different house each evening but the days are mostly free now. My great niece (we've been working on her calculus homework some evenings too) told me she was thinking about a double major. I told her I didn't think I could handle a double major. She got a kick out of it. Actually I've been a lot more help with the theoretical, explain why equations fall apart here or there than the mechanics, she blitzes through the calc and algebra so fast there's always a pregnant pause before I my mind catches up to nod my approval.
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