Single link front end design tool WIP - feature requests?

Discussion in 'Some Assembly Required' started by sebwiers, Apr 21, 2018.

  1. sebwiers

    sebwiers  

    Joined:
    Aug 2, 2016
    Oddometer:
    91
    Location:
    Saint Paul, MN, USA
    As an extension of my chopper planner / front end configurator tool, I'm working on a version that calculates the wheel path (and some resulting kinematics) for a springer style front end (would also work for Erles fronts, and whatever those awful schwin bicycle style front ends are called - basically anything with just one pivot).

    Is this vaguely useful to anybody? What sort of calculations are of interest?

    To my mind, anti-dive is the big one. Chopper style springers (if they have a front brake) usually run some sort of floating brake caliper, which makes calculating the dive force much more complex; how common is this on other setups? Is it worth the extra work?

    I almost certainly will include options for at least basic spring setups, to calculate wheel rate etc.
    #1
  2. larryboy

    larryboy Chopper Rider

    Joined:
    Jan 9, 2005
    Oddometer:
    20,752
    Location:
    Über Alles,California
    I'd be interested. I've done a leading link on a chopper and just made some educated guesses that worked out pretty good, but knowing beforehand would be nice.

    I'm thinking about doing the Dirtbag Challenge this year and I'd do another leading link front end.

    :lurk
    #2
    Roadracer_Al likes this.
  3. sebwiers

    sebwiers  

    Joined:
    Aug 2, 2016
    Oddometer:
    91
    Location:
    Saint Paul, MN, USA
    Is it worth working out anti-dive with fixed brakes, given the popularity of floating brakes in that application? Changes to static geometry are MUCH easier to work out - in fact I've mostly nailed them after one weekend work, just need to work out input (its running off some hard coded test values) and graphing at this point to have those. Might be a bit janky at first, but I should have a minimal working version up in a week or less.
    #3
  4. larryboy

    larryboy Chopper Rider

    Joined:
    Jan 9, 2005
    Oddometer:
    20,752
    Location:
    Über Alles,California
    The one I built had a fixed caliper.
    #4
    sebwiers likes this.
  5. Dino de Laurentiis

    Dino de Laurentiis Working on it

    Joined:
    Mar 20, 2010
    Oddometer:
    239
    Location:
    Ultima Thule
    When you say floating caliper i assume you mean something else than the typical mounted on sliding pins floating caliper? If not, how does the fact that it’s sliding affect anti dive?
    #5
  6. larryboy

    larryboy Chopper Rider

    Joined:
    Jan 9, 2005
    Oddometer:
    20,752
    Location:
    Über Alles,California
    Floating:

    646442380_GAgyj-L.jpg

    Fixed:

    P1070994_zps58709bd4.jpg

    I don't know the answer to the anti-dive, hence the interest in a calculator.
    #6
    sebwiers and High Octane like this.
  7. gwarden

    gwarden Adventurer

    Joined:
    Feb 27, 2018
    Oddometer:
    33
    Location:
    Southern CA
    I'd recommend drawing up a schematic in a program like Linkage to analyze anti dive and more importantly leverage rate.

    Algodoo is also a fun, if somewhat cartoony, way to design linkages. I know a few Ultra-4 cars who had linkage designs originate in Algodoo.
    #7
  8. sebwiers

    sebwiers  

    Joined:
    Aug 2, 2016
    Oddometer:
    91
    Location:
    Saint Paul, MN, USA
    I am drawing up a schematic. I'm just using common web browser programming to do it, instead of a commercial program. Calculating all the numbers myself also allows me to dump them into a popular graphing tool which also draws on a web page.

    I think I tried Linkage (or maybe something like it) once and couldn't figure out how to make it go. I'm trying to make something that is MUCH easier to use. It might be limited to a range of special cases (single pivots with fixed brakes or whatever) but I don't mind working my way through a series of special cases, since I'm mainly doing this to hone my skills as a programmer. I designed my bike using Tony's software, and while it has some annoyances I hope to improve on, it does a great job. The main downside is it costs $75 and you can't just download it and start using it (because his rights management depends on direct email) which discourages people from using it to play "what if", and also prevents people from doing analysis on commercial offerings to see if they are shit or not.

    By putting something free on the web that is easy to use, I'm hoping to encourage curiosity about the designs that maybe results in a few more interesting bikes getting built. I might try packing it all up as a phone app eventually, but building it as a web page first means I can literally just dump my code into a locally stored app (it doesn't need a web connection to work, it's all running inside the browser).
    #8
  9. sebwiers

    sebwiers  

    Joined:
    Aug 2, 2016
    Oddometer:
    91
    Location:
    Saint Paul, MN, USA
    I mean the sort where the caliper mount can rotate about the axle, and is prevented from doing so with an attachment to some part that moves differently from the axle (usually a part that move not at all, like the fork leg).

    Its not an idea unique to pivot forks, its just more common to them. Here's an example on a telescopic fork. I think its fairly obvious that braking force is actually used to push UP on the fork leg, countering brake dive.

    [​IMG]

    It's pretty simple for cases where the caliper is fixed relative to the axle and the bike remains level (which I'm assuming it does; bumps come in all different shapes, so I'm picking the one that makes calculations easiest). This video explains the methods better than anything else I've seen (outside a book anyhow)- he's talking about carss, but the methods are exactly the same when talking about motorcycles, you just use different bits to control the wheel path / instant center. For a single pivot, the pivot is always the instant center...



    I could probably muddle through the additional effects for floating calipers using half remembered high school physics, but computer programs don't allow muddling. I'll probably have to do a few cases on paper, and see if I can work out a way to do it in code that is reasonably efficient. Probably the sane thing to do would be to do the "simple" case of a sliding fork like in the above picture, which is actually something I've wondered about anyhow...

    Anyhow, I've got the basics working now, gonna hok it into the graphing package and throw it up on the website. For now its just trail / wheelbase vs bump travel, anti-dive will be a new ball of wax.
    #9
    larryboy likes this.
  10. sebwiers

    sebwiers  

    Joined:
    Aug 2, 2016
    Oddometer:
    91
    Location:
    Saint Paul, MN, USA
    #10
    larryboy likes this.
  11. gwarden

    gwarden Adventurer

    Joined:
    Feb 27, 2018
    Oddometer:
    33
    Location:
    Southern CA
    @sebwiers nice visualization, looks fantastic and I may make something similar for a course I am teaching this fall.

    One minor thing to contribute, though: I may be reading it wrong, but I think you may want to tweak around a bit so that you have an increase in trail as you go through the travel. The Telelever system picks up multiple inches of trail, though I don't have a graph of trail vs. travel. To my memory the GSA picks up > 2" of trail over 7.5" of wheel travel.

    If you have MATLAB I can send you a script to calculate stability vs speed.

    I'd also be curious to see your leverage curve, though it doesn't look like it will be a problem.

    Loving this project!
    #11
    sebwiers likes this.
  12. sebwiers

    sebwiers  

    Joined:
    Aug 2, 2016
    Oddometer:
    91
    Location:
    Saint Paul, MN, USA
    @gwarden Yeah, I definitely messed up the math for the axle motion calculations! In certain cases you can get the axle to track a full circle that isn't anywhere near the pivot, so its obviously rotating around some location that is NOT the pivot point, but was close enough in my hard coded test case to that I did not notice. Now that I can try multiple examples from page input, I'll get that fixed pretty quick. I'm probably mixing up +/- or sine/cosine.

    Good luck with your class / projects! I don't have matlab, and am wondering - how even CAN you calculate stability vs speed? Is it a calculation of the straightening force on the bars due to trail & rolling resistance, or is there more to it than that? I'm trying to avoid multi-variable calculations, both because they are hard to code, and because people often have no idea what the figures mean, and they may be based on certain assumptions. Heck, even in this case I am making a big assumption when calculating trail, namely that the road surface under the front wheel is level while the wheel moves up and down, without causing any frame pitching. In the real world, trail changes all over the place as you go over a bump due to the wheel contact the road ahead of / behind the point directly under the axle, the rake changes as the frame pitches, etc.
    #12
  13. gwarden

    gwarden Adventurer

    Joined:
    Feb 27, 2018
    Oddometer:
    33
    Location:
    Southern CA
    @sebwiers gotcha, can't imagine how much time we as a race have spent staring at a screen/paper thinking, "I think there is some trig that's backwards here..."

    The stability vs. speed calculation is a bit of a bear, which is why the MATLAB program is very handy. Oh well - I'm sure you'll have a very stable setup as long as you don't get close to zero trail.

    If you're curious about the stability problem, it's an eigenvalue of a large matrix. Here's a great paper that I use as a handout. There's a lot going on and admittedly it's a little busy, but if you look at it for long enough it starts to make more sense.
    https://www.princeton.edu/ssp/64-tiger-cub-1/64-data/stability-of-motorcycles.pdf

    The key takeaway from all that is really Figure 5, which is what the MATLAB program spits out. You're exactly right, the result is essentially like a straightening force at the bars, but it's all a result of the inertia of the wheel, center of mass, etc. Technically it is a damping coefficient that allows you to find stability. You can use the damping (which is given as an equivalent torque on the bars) to figure out at which speeds the bike is most likely to either tip over of go out of control based on the natural frequency of the system.

    When looking at the graphs, a negative value for the damping value is a tendency for the bars/bike to straighten out, whereas a positive value would suggest the bike will fall right over. Fun stuff.

    Are you planning on having a shock offset from the wheel or directly in line with it?
    #13
  14. sebwiers

    sebwiers  

    Joined:
    Aug 2, 2016
    Oddometer:
    91
    Location:
    Saint Paul, MN, USA
    I fixed the travel calculations.

    I also tweaked things so that it handles both leading and trailing arms gracefully, recognizes that negative trail is a thing (before it was just calculated as a distance), calculates a minimum of 100 points in travel (so that short distance travel graphs aren't sparse on points), and doesn't go into an infinite loop when the entered bump travel is greater than the link length allows (though the resulting image and graph are a bit wonky).

    I also altered the visualization so that it draws the trail lines along with the pivot positions that correspond to them in a matching range of colors. Sounds confusing, is pretty obvious when you see it. Another tweak is it draws as soon as the page opens, and draws the chart and diagram with one button.

    https://sebwiers.github.io/motorcycle-front-end/sandbox/single-link.html

    [​IMG]

    @gwarden - I don't think I'll be getting anything from that paper, the math is way beyond me. I have watched a few videos about bicycle stability that showed me there's more to it than trail & gyroscopes - stuff like how the COG lifts / falls as the wheel turns, the wheels COG relative to the steering axis, and other oddities. Interesting, but beyond my pay grade. Not sure what you mean about the shock... you mean for something I'm building? I already built my bike, not planning another anytime soon. For the software, any shock calculations I do will have to be flexible enough to allow arbitrary positioning of both shock ends. The bike setup coded into the software isn't anything I'm building, its just a setup that I could work out some numbers for in my head, that I used as a test case.
    #14
    gwarden likes this.
  15. Roadracer_Al

    Roadracer_Al louder, louder, louder!

    Joined:
    Aug 5, 2007
    Oddometer:
    2,216
    Location:
    Oakland, CA
    Subbed!
    #15
  16. sebwiers

    sebwiers  

    Joined:
    Aug 2, 2016
    Oddometer:
    91
    Location:
    Saint Paul, MN, USA
    Update. Cleaned up some small annoyances, improved drawing, filled in the missing numbers in the table, created a link on the index page. Not sure I love the rainbow effect on the drawing, but it goes the full way around the color spectrum for any given data set so is more consistent.

    https://sebwiers.github.io/motorcycle-front-end/

    [​IMG]
    #16
    Roadracer_Al and larryboy like this.