Time to see if the steering adjustments are working…
…and the brakes too 🙂
Time to see if the steering adjustments are working…
…and the brakes too 🙂
Once I knew I had a winning ticket for the race, I did what any programmer would do when getting started.
No, I didn’t make the t-shirt. I wrote a simulation.
Step #1 was to get the profile of the race course. Google Earth has nice support for generating a profile from a path:
Next was figuring out how to simulate the race. For every meter down the course, the profile lets me calculate the change in elevation, and thus the change in potential energy. Once I have this, and the mass of the vehicle, and the resistance due to wheels and air drag, and the moment of inertia of the wheels, then it was a simple matter of mapping from delta PE to delta KE:
And then I could turn these equations into code:
Finally, I could generate results with different weights, and graph them.
The sad result was that more weight == better results. Though I imagine rolling resistance isn’t linear with weight, as the “footprint” of the tread expands under increased load. Plus there’s the issue of being able to stop at the bottom of the hill…
I’ve been keeping a collection of emails sent by my Dad, as he works in his shop to get a disk brake attached to our (small) wheels. I think it’s pretty amazing that he’s got the energy and persistence (at 85!) to keep working away at a problem until he solves it.
Two days ago:
I have been chasing my tail for several days trying to get the brake discs to rotate without ‘runout’ when installed with the wheel on my support brackets. I have gotten as close as +/- .005″; not good enough.
Each time I lapped the face the disc bolts to, and reinstalled the disc, I got different results. I tried adding shims to the bearing outer races – no luck.
Finally, tonight I removed one of the spacers from between the bearings and measured the length at 4 points with a micrometer. 1.0303, 1.0302, 1.0295, 1.0288. That calculates to an angle of 0.1836 degrees.
Now I wonder how to true up the sleeve. I don’t know which end is off. And I don’t know if my lathe and chuck are accurate enough. I can also chuck it in my milling chuck. If I grind off a little I will need to add a hard stainless shim type washer to compensate and keep the length to 1.030. I have shim stock in various thicknesses.
1) I checked out the lathe; it is pretty accurate.
2) I trued up the sleeves by chucking in the lathe and grinding with my Dremel. One end was off on each of them.
3) Brake disc run-out, and the wheel still wobbles!!
And today, victory!
Today it finally dawned on me that the top hat we made to mount the brake disc to the wheel really forms a backbone for both. The disc is pretty flexible; the wheels are made from fiberglass reinforced plastic which has a low modulus of elasticity. In between is the rigid metal top hat.
By adjusting the bolts to the wheel, I have the one wheel running less than +/- .002 instead of +/- .020. Now I am working on the face of the hat against the disc. If need be I can shim and get it accurate. The gap in the brake clamp is .090 with a little drag on the feeler gauge. The disc is .070 thick, leaving less than .020 total.
I’m glad he’s on our team 🙂
The Nevada City Adult Soapbox Derby event is approaching with frightening speed. Time to increase the development velocity of our Tube of Terror.
And it’s looking smooth and true…nice.
Though maybe not as aerodynamic as we’d want.
Back in 1985, I was a young programmer at Apple who wound up spending time in Japan, helping Apple evaluate options for supporting Japanese on the Mac.
This nascent project became the focus for Apple’s new Pacific division, and a manager was dutifully assigned. A former sales guy named Dave Kleinberg. Great. Just what I always wanted.
And yes, there were some early impedance miss-matches, but by the end of the project he’d earned my respect. There were countless details outside the scope of just “gettin’ er done” (the coding bit that I cared about), and Dave sweated the details. We wound up shipping KanjiTalk 1.0 in May of 1986, and this wound up being the foundation for Apple’s long term success in the Japanese market.
As a side benefit, Dave gave all team members the best project tchotchke ever – the KanjiTalk Monolith:
Why the post today? A member of the KanjiTalk team just sent me the link to Dave’s obituary.
Dead at age 53 from lung cancer. We’d seen him a year ago at the 25th anniversary get-together, and he’d seemed fine. I wish I’d told him then what I just wrote now.
I grew up in southern California (Whittier), which is pretty close to Knott’s Berry Farm.
Back in the 1960’s, this wasn’t yet a typical theme park filled with roller coaster rides. There was a nice old-time feel to the place, with chickens and peacocks wandering around the parking lot.
We’d visit regularly, and every such trip included a viewing of The Button Collection.
My grandmother lived in Filer, Idaho and collected thousands of buttons over the years. Eventually an aunt finished off the collection and donated it to Knott’s Berry Farm, where it’s still on display.
June 16th is the second annual Nevada City Adult Soapbox Derby, where grown men (and some women) raise money for Pioneer Park, by spending ridiculous amounts of time building gravity-powered vehicles.
Here’s the local hospital entry from last year:
This year, since I’ve got so much free time, I decided to put together a team. Thus was born the “Tube of Terror”, sponsored by Scale Unlimited. With the aid of some people who actually know how how to design and build things, I’m hoping we’ll be competitive in the speed category.
Almost four years ago I’d written a blog post about the members of my MIT fraternity class – Whatever Happened to the Class of ’83
I thought I’d revisit that theme, but now with an updated class composite photo, using images taken during a recent 50th birthday event hosted by Jono Goldstein at his place in Cape Cod.
Updated details to follow, when I get some time…
Everybody knows about the Linux “epoch”, as in January 1st, 1970 UTC.
Some people might remember the original Mac OS equivalent, which was January 1st, 1904. From what I remember back in the day, Jerome Coonen picked that date (instead of 1900) because it simplified the leap year calculations. And using an unsigned 32 bit value for seconds meant that Macs could handle dates up to 2023 or some point way in the future.
Then yesterday I came across the MUMPS/Caché “$h” date format, which is described as:
This format returns the date as the number of days since 1st January 1841, and the time as the number of seconds since midnight.
I was speculating with my friend as to why they picked 1841, and guessed that this was some convenient date (similar to the Mac 1904 choice) that was before the birthdate of the oldest person they could imagine being in the system, back in the 1960s.
After turning to the Source of Truth (Wikipedia’s article on MUMPS), it seems like our guess was correct. James Poitras explains why he picked this odd date:
I remembered reading of the oldest (one of the oldest?) U.S. citizen, a Civil War veteran, who was 121 years old at the time. Since I wanted to be able to represent dates in a Julian-type form so that age could be easily calculated and to be able to represent any birth date in the numeric range selected, I decided that a starting date in the early 1840s would be ‘safe.’ Since my algorithm worked most logically when every fourth year was a leap year, the first year was taken as 1841. The zero point was then December 31, 1840