Iterative design

I use a design process called “Iterative Design”. That is I make a design, then go back and look for mistakes before I freeze the design, or I make major changes and do a comparison to see which is better, then combine the best parts of both for the next iteration. This process grew out of “Trial and Error” which is an offshoot of “Cut And Try”. Anyway, that is the major reason why I take a completed design and make major changes or even start completely from scratch.

Of course some of the times I do that it’s because the design parameters were changed by factors beyond my control, as when the minivan donor vehicle became unavailable. That was an entire design that had to be scrapped, and started from scratch on the Pentastar lightweight replacement design. But the thing is I’m doing this because this is how I do things. Every so often I just click on “Start a new drawing” and start drawing a new design. Part of that is because I know I’ll never have access to the money or materials I need to build it, part is because if I stop creating I get antsy, and other psychological “issues” crop up. Fortunately I have two outlets for creating, this blog and planning the Sprint-T.

I have also been thinking about transcribing some of my poetry to the blog. Most of it is about 20 years old at this point, and the Y2K point-and-laugh in particular are a little dated (like Conan’s “Year Two Thousand” bits). But some are still pretty good if morbid, like “Musings On a Bird, Skeletonized By Ants”. Tell me what you think in the comments.

And it’s getting early in the AM and I have things to do tomorrow.

Advertisement

Wow was I underestimating my walks

Someone mentioned an app on Twitter that did something I needed, tracking walks and estimating calories burned. So I looked it up at the Google Store and found out the price was right (free) and the basic version did in fact do everything I needed from such an app. So I now have “Map My Walk” installed on my phone.

What I learned is my “short walk” is 3.31 miles long instead of my estimated 2.5 miles I’ve been telling everyone, and that I burned 500 calories on my “short walk”. I’m thinking my “long walk” is closer to 5 or 6 miles than the 3.5 or 4 that I thought. This also explains how my shoes keep wearing out so fast, since I go for a walk almost every day. I’m also faster than I thought but still nothing to brag about.

Also while on my walk I contemplated the one remaining issue I have to solve with the rear suspension: keeping the rear wheels from steering the car under power because the toe angle changes. With the 2D Pratt truss the knuckle could have had enough bending torque to change the toe angle uncontrollably, so I had to make it 3D without adding to the weight or making it take up too much space behind or over the transmission. Thinking the problem through I realized the 4-link was going to be taking almost all of the forces that would be changing the toe, so all I really needed to do was give the 2D truss a fighting chance at using the depth of the structural members to resist toe change by using some diagonals connecting to the vertical member of the last bay before the knuckle, in as many planes as I could so the last bay would be 3D but everything else could remain 2D, only slightly increasing the weight of the truss and not changing the vertical stiffness needed to keep the wheel aligned on the other 2 axis. In doing so I would not be using a true Pratt truss in that the diagonal members will be under compression rather than tension, and the end bay would not have an upper horizontal member or terminating vertical member but would be similar to a kingpin truss in 3 planes, with another triangular bay under the truss from the bottom of the knuckle up to the bottom horizontal member in the plane of the main truss keeping the camber angle constant. I wish my CAD skills were good enough to draw this out and output a .gif or .jpg file to put in the blog, because words are only so good in describing this.

Another stab at it, the plate that holds the knuckle to the truss is going to be rectangular or maybe a right triangle with the hypotenuse facing down and forward and the right angle on the top rear side. Diagonal members will run from all 3 corners of the triangle to the first vertical member of the truss, one from the bottom corner to the bottom of the first vertical member from the bottom, one from the top front corner to the bottom of the vertical member on its front side, another from the top front to the top of the vertical member on its front side, and a last one from the top rear corner to the top of the first vertical member, that diagonal in the plane of the truss. That should be more than enough to keep the forces driving the wheel under control so the toe and camber don’t change. And none of the added members will interfere with the drive shafts or the lateral links of the Watts link that keeps the whole shebang located from side to side. That’s because all of the added structure is either behind or above the knuckle and therefore above or behind the driveshafts and lateral links.

I really need to learn how to use the AutoCAD 360 program I downloaded to the laptop and not just so I can draw pictures of the car to post to the blog. Anyone know of some good tutorials to teach me to use it?

Billed @€0.02, Opus the Unkillable Badass

More on that T-bucket

I know my old readers from back in the day 8 years ago today (yay!) might be getting a little tired of my building a car here. but I have to write and I can’t write about bicycles for a while. So building a car from scratch and scraps it is for now. Are any of you reading now still reading from back then? Leave a comment please if you are.

I have pondered the plusses and minuses of the three suspension designs I mentioned a couple of days back and made a decision based on what will work best and still not break the budget. Before I tell the winner I have to explain what went into the decision.

The overall design of the original car called for a tube front axle with reproduction early Ford spindles running adapted Mopar disks and GM Metric calipers held up by coilovers and connected to the frame with a parallel 4 bar and a long panhard bar for lateral location much like a sprint car. Since I had already bought much of this setup economics demanded that I make the rear suspension compatible with this. The biggest feature with this suspension is zero camber change in roll compared to the road as long as both tires remain in contact with the road, and a mostly fixed roll center height. This makes for a predictable and controllable front end which is worth way more than extracting that last 0.01 lateral g from the front tires. But to make the front suspension work the rear suspension has to be equally predictable and controllable.

So we start with the minivan strut suspension. Plusses are mainly low cost as all the parts are right there. Most of the minuses were spelled out in that other post, the biggest being lack of tunability , and a secondary one being that ride height would have to be set during the frame building process without knowing the spring rate or how much the back end will weigh. Well in addition to the construction issues there are the geometry issues. The biggest of those is the camber changes pretty much in tandem with the body roll, causing the outside tire to roll over on the outside shoulder and pretty much killing mechanical grip making the back end much less predictable and also drastically reducing overall grip. So as much as it pains my Scots ancestry this cheap solution ends up not the correct solution.

The second possibility was using a heavy wall aluminum tube like the axle tubes on a quick change rear end for a front engine car to make a de Dion rear suspension. This has pretty much the same geometry as the front end making it a much better choice for the midengine Bucket than the salvaged struts from the minivan. It is very easy to make the camber and toe adjustable at the plate that mounts the minivan knuckle to the tube which aids in setting tracking and adjusting the rear to match the front. Another advantage is the coilovers can be located pretty much anywhere on the top of the tube making setting the initial roll stiffness very easy. The major disadvantage is I don’t have the equipment to weld aluminum except in very thin material so the welding would have to be farmed out. That is an additional expense.

That brings us to the third possibility, the steel truss de Dion rear suspension. Again we have a very compatible geometry for the front end, that is also highly adjustable for everything and has a similar weight to the aluminum tube version, but one that I can build using the equipment I have on hand. One slight disadvantage is the mounts for the coilovers have to be over or very close to a vertical member of the truss (I decided on a modified Pratt truss) which means either the truss will have to be iterated somehow (probably with a shear plate that would be removed after the mounting point was determined) or the initial mounting point will have to be SWAGed (Scientific Wild Ass Guess) and then tuned with an anti-roll bar. For simplicity I would prefer to not have anything extra on the car for street use, so SWAG for the initial mounting point on a shear plated truss then do a limited iteration on both sides for the final mounting point, then complete the truss. Or just leave the extra mounts and one bay of the truss shear plated for adjusting the rear roll stiffness at the track. I like this because instead of bringing a rack of anti-roll bars to the track I just move the coilover from side to side on the axle to adjust the balance of the car.

So the Final Solution to this problem is the steel tube Pratt truss with one bay shear plated and several mounts for the coilover along that bay and matching arrays of upper mounts on the frame crossmember on the rear, with the tube axle and early Ford spindles and the coilover mount as close to the spindle as possible without getting tire rub, parallel four bar longitudinal location on both ends with a long panhard rod on the front to eliminate bump steer from the cross steer steering and a Watts link mounted low in the center of the frame running out to the knuckle mounting plate for lateral location. And I literally came to this conclusion while composing this post, because that’s how I think. You just got to watch my mind at work iterating a design specification. I had already decided on the truss but the details got finalized while doing the post.

Billed @€0.02, Opus the Unkillable Badass

RPG night was a bit chaotic and other stuff on Wreck-Free Sunday

Yesterday was RPG night, with a bit of a fluster cluck to start out. We had to move to a different venue when both of the hosts at our regular location were called in to work, which really screwed the meal planning. Our regular location we have access to a kitchen with normal pots, pans , and a slow cooker, the backup location we don’t have kitchen access so the ramen and beans menu planning went out the window. I got some cash from Mrs. the Poet for transportation and food and I got some “mexican” food from Taco Bell.

I had to create a new character because the character sheet for the one I just finished creating (the F.I.S.S) was at the regular location, and I had an idea for a bicycle-based speedster. My new character has the superhero name of “Flash Drive” and wears what looks like motorcycle leathers and helmet with a full-face shield to hide his secret identity as Clyde Bonney, international ultra model. Yes, going back to his origin story, Clyde Bonney was just a better than average bicycle messenger when he was sandwiched between two cars running the same red light in opposite directions. The Gods intervened with the single comment “Well, that sucks,” and granted him super speed (288 MPH sprint and extended cruise) and super toughness, genius level thinking and deduction powers (think Batman), and since there wasn’t much left of his original face they rebuilt him with god-like levels of male beauty, all in an instant. He went into the intersection a high-level normal human and got squished between two guys running that red light, and reappeared without his bike next to the wreck still moving as fast as he was before the wreck but on foot instead of a bicycle, and he literally walked away from the wreck with the remains of the bike trapped between the two cars. The wreck and subsequent publicity resulted in a career in international modelling and high society with a high middle-class income and lots of “perks” like free clothes and sometimes the use of cars. This results in the disadvantage of “the curse of Beauty” in that almost all bi people of both genders, and gay men or straight women, are pursuing him for a relationship or just sex. I have already decided that when it’s time to replace my costume with something a little more bulletproof my secret identity is going to “Cosplay” my superhero identity in the old costume, and model the cosmetic proofs of the new costume in a national “America chooses the superhero’s new costume” publicity campaign. Mind you this is in a universe that has a superhero-only edition of “People Weekly”.

The first adventure was getting the party together, as we have a retired ninja, a genetic super-soldier werewolf type character, and my speedster character working as part of a team in the small New England city of Cape Temperance. The big problem we have is we have two characters with rather high caloric needs, mine and the werewolf that looks like Bigfoot in his human form, and the werewolf does not know who he is because he was getting implanted memories during his “vat stage” of development and that part was interrupted by the lab he was growing in being raided in the last game session which we played with villain characters. There was a lot of RP in the session with only a single combat with a dumb “brick” character who was strong and tough but not very bright. My character managed to get some blunt force damage in, and the ninja was able to significantly slow him down with his sword, with the werewolf character keeping the brick’s hands busy until he managed to take the brick out with a clawed kick to the throat that brought him down just as the media showed up. Flash Drive was given partial credit as the ninja was invisible most of the time especially when he would attack with his sword someplace I had just attacked visibly with my fists at super speed.

As I tell Mrs. the Poet I role play because it’s fun and therapeutic. I get completely out of my normal headspace as a bicycle advocate and journalist and into the headspace of my character, which makes the time I have to be “just me” much more tolerable. I find the best way to cope with life is just to leave it behind for a while and get out of my normal way of thinking.

I finally got the video game I asked for for Father’s Day, Gran Turismo 4. Now I need to get a controller that will let me control my car better. Basically I have the choice of another standard controller or one of two Logitech steering wheels, the Driving Force Pro, and the GT Force. I tried the standard controller with the analog stick under my left thumb controlling the steering and no matter what car I was driving I was all over the road in it. I was able to keep a car on the track at the Motegi Super Speedway, but only just barely by putting hard compound tires on the front and super soft on the rear, causing a huge push that allowed me to be able to move the stick a visible amount. Setting up a car like that on any other track would result in plowing off the track. So, before I can build the virtual version of my car, I have to find a controller I can drive with, and then learn how to drive in the simulation, then create a virtual copy of the Sprint T. And this is what I’m starting from on that.

That last picture is really good to work from the CAD drawing (2D) perspective as it is almost a square-on with minimal perspective view, I could trace that and have a very good starting point for revamping it to the Sprint T configuration. Imagine that body with the cage from this car slapped over and around it.

Note the tubes running from the top of the roll cage to the front of the car, they help with triangulation of the frame and they increase the depth (d) of the frame, basically making the upper part of the roll cage the top rail of the frame. With the bending moment being a function of d⁴ everything that can be done to make the top of the cage function like the top frame rail has massive effects on the stiffness of the frame, and the more space I can make the frame enclose just does the same thing to the torsional stiffness of the entire car. In simpler terms moving the diagonal brace from the top of the front hoop so that it connects to the top of the front crossmember and then running another brace from the crossmember to the bottom of the front hoop makes the car hella stiff in bending and torsion. Running diagonal braces inside the cage to the bottom rail from the tops of both hoops makes the center bay of the frame just that much stiffer again, and gives me more to grab hold of to get in and out of the car. Putting the cage (mostly) on the outside of the body does the same thing, making it stiffer by making it enclose more volume while keeping it triangulated, and by doing that I can use smaller and lighter tubes in areas not specified by racing safety regulations, which makes the whole car lighter. The Speedway Motors kit uses a 1½ by 3″ by 0.120 wall rectangular tube as the sole frame member for their kits.

But the Sprint T could use the 1½” by 0.060 wall square tube as the bottom rail which weighs about ¼ of the Speedway Motors part. The Sprint T frame will weigh slightly more than a standard T-Bucket frame, but the difference in torsional stiffness would be incomparable. Think heavy ladder compared to light jungle gym. Compare the number of tubes and space enclosed on this frame to the one on the Speedway kit.

The roll cage hoops and internal diagonal braces have to be 1½” by 0.120 wall chrome-moly round tubing because those are required to run an open car in SCCA Solo competition in the Modified category which is where the Sprint T will wind up, but aside from the 1½” by 0.060 wall square tube bottom rail most of the frame will be 1″ by 0.060 wall round or square steel or chrome-moly (depending on what is available when I go shopping).

And I got called old politely again last night on the return trip from the RP game session. There were some rowdy kids that got on the train drunk (yes I said drunk kids) and one of them had problems not sitting on me when he was trying to talk to the rest of his party. He wasn’t being malicious or even obnoxious, but it was slightly uncomfortable. Another person on the train admonished the kid to be more respectful because “what if he was your grand-daddy?” :scowl: Dangit, I’m not old enough to be a “revered elder” yet. 😉

And that’s all I have to say today. Sorry about the squirrel chase around the topics. 😉 😛

PSA, Opus