Injection Moulding and Thinking Ahead

Have a quick look at this picture. What do you see? If your answer is "an injection moulded front panel of some sort" your answer is correct. So why this image? Because it illustrates two things, one of which is very important, the other actually being a minor thing.

What this illustrates is the importance of thinking ahead when designing your components. This part wasn't originally done by me, it was done by a colleague of mine. He did an ok job with it, basically doing what he could to replicate an existing part. But he missed a couple of crucial things. This part is injection moulded, so draft angles are required. We normally go for 0,5 degrees on relatively short surfaces and wherever there is contact between parts of the tool we go for somewhere in the range of 3-5 degrees. The draft angles at the openings, which clearly need to be 3-5 degrees, were not done when I received the model. Also, we need to be as structured as possible when designing parts, striving to keep it so that it's easy to see in the model tree what has been done and avoiding using an excessive number of operations (e.g. let's say we have a number of screw bosses with reinforcement ribs and a few fillets, then we would make one screw boss complete with ribs and fillets and pattern it as a whole instead of making each boss as a separate entity). This model was not easy to understand.

As I was given the model to finish it up to send to the intended supplier for evaluation I found a couple of issues. Not only did it have a lot of operations which could have been avoided but as I tried adding the 5 degree draft angle at the openings, the software protested. The shape was simply too advanced for the software to accept. Catia will sometimes protest to things that seem very simple. The solution was to go back to where the cutouts were created and re-do the surfaces defining the cutout. Ok, fine, that gave me the required angle. But then the shell operation wouldn't agree. In the end, I redid almost the entire model.
This could have been avoided had my colleague simply thought of the required draft angle around the openings much sooner and strived to keep the model more structured. It's a mistake we all make at some point, forgetting about something specific to manufacturing or not keeping models structured, but it really shows the importance of thinking ahead when designing stuff.

What is that other thing that this image illustrates? The importance of having an easy-access tool for snapshots. I know Catia has one, I've used it many times, but I have not been able to find this kind of tool in Inventor. I found a way to create the image but it was by no means quick and it did not produce the image I actually wanted. I could of course have used the print screen option and then cropped the image to include only that which I wanted to show but that should not be needed now should it?
Of course, I could have missed something in which case you may feel free to educate me further.

The rear suspension is getting somewhere

Last time the rear suspension was giving me headaches, it was very frustrating. Well, it turned out that some of that was due to constraints initially set between one of the control arms and the upright. What I did now was to make a skeleton for the suspension, i.e. I inserted a part into the product which defines the pivot points on the chassis. This way I managed to get a rough estimation of where the pivots should be. Which means it's now back to measuring the drawings and figuring out where the suspension should attach to the frame... I also need to examine the suspension and decide what kind of joint should be used where, it doesn't seem appropriate to use only ball-joints but maybe I'm missing something.

Pictures speak a thousand words, so here's what I have for you today:

Rear suspension in progress

Yet another critical item to get right in a racecar is the suspension. So I had better model it, hadn't I? I have concentrated on the rear first simply because I found it more interesting

It seems I will need to look for bigger wheels...

The rear knuckle

I can't remember if I ever put pictures up of the rear knuckle I modelled before getting the deal to model the IMSA spaceframe for Paul. In case I didn't post them, here they are

Brake upgrade update

Since the last post I have updated my model and also actually used my upgrade on the street.

I don't remember if I told you in the previous post but the front face of the knuckle wasn't plane so I decided to mount the adapter to the rear face just like the original caliper bracket. I then used spacers to put the caliper in the correct position axially.

So it all worked just fine until I first managed to put a bit of heat into the rotors after a high-speed run. Building up enough heat took more than I thought, but as I did there was a noticeable interference happening between the caliper bracket and the rotor. Looking through the wheel the result of the interference was most obvious for the left-hand rotor and upon inspection I noticed I had radial interference (see highlighted surfaces in picture below).

I'll have to take care of this. Since I seem to have a margin of 5mm between the pads and the outer edge of the rotor I could redesign the bracket to move the caliper 2mm radially but I also want to know what is wrong with my model, especially considering the assembly shows my design should already give more clearance than the classic bubbajoe adapter.

IMSA update

Not a lot has happened since my last post, but it hasn't been standing still either. As I wrote earlier, the chassis stiffness was up at 8139 Nm/deg in my FE analysis. Not too bad, but really not excellent either.

So, what improvements have I found? Well, there were limits to what could be done. Paul figured that adding tubes in the floor on the driver side would get in the way of his seat and pedals. I agree that it might result in a higher CGH. It's all a compromise and in the end, it is Pauls car and his word is law.

But I did add a brace to the floor on the passenger side. If Paul was to have a passenger with him, the extra weight of the passenger would be more of a problem than the slightly higer CGH anyway.
With the floor brace, the stiffness was now 9811 Nm/deg.

I then looked at further triangulation of the passenger compartment but didn't get higher than 10313 Nm/deg and I considered this to be a pretty small gain. It just didn't seem to be worth it.

Having studied the deformed mesh, I concentrated on the rear end. I added some braces connecting the engine mounts and the square section tubing at the rear. This time, the results were a bit more interesting: 12649 Nm/deg.

Now, while this was a nice improvement, I am aware that these braces could come in conflict with the driveshafts. Unfortunately, Pauls attention has turned towards a different project at the moment, so he hasn't been able to check this yet. Until he has checked this, or come up with another idea on chassis improvements, I guess there's not much more I can do right now.

Here are a couple of pictures of the frame in its current form

Enjoy

PK

Spaceframe comparison

 

Finally it seems I have the spaceframes modelled up. Above is a picture of the original IMSA frame

And here is a picture of Pauls spaceframe. I know the frames aren't exact, as an example they lack suspension pickup points. But they're both close enough for now. Further refinement will be made.

 The point was to compare the torsional rigidity of the frames and then work to find improvements for Paul. The picture above shows how I set the analysis up. A virtual part was attached to the front rectangular beam which will house the attachment points for the front shock absorber. This part was then rotated 1 degree around the longitudinal axis. At the rear rectangular beam, another virtual part was applied and this was clamped.
I do know this is not really the correct way to do it if you want values to compare to the real world, but for comparing spaceframes and see if improvements can be made I felt it would be good enough, especially since there's a limit to what I can actually do on my computer.

For Pauls spaceframe, we suspected that the engine would act as a cross brace. I decided to model this by adding an infinitely stiff virtual part connecting the engine mounts.

For the analysis, a full-on 3D analysis seemed appropriate. To speed up the computations, the mesh size was set to 20mm with a sag of 10mm using parabolic elements. Material was set to steel.
And the results?
For the IMSA spaceframe pictured at the beginning of the post, a moment of 6168Nm was recorded.
For Pauls spaceframe, not including the effect of the engine/transmission, the moment was 6824Nm.
When the transmission was taken into account, the moment increased to 8139Nm.

I am confident that we can find ways of increasing torsional rigidity for Pauls car. As a reference, speedlab have their corvette racecar up at 18000Nm/deg

Now, next on the to do list is finding improvements

Hosler's IMSA

What you see in the pictures above is a wet dream of many Fiero enthusiasts. This is an IMSA Fiero, a real one, not another bodykit on yet another "ordinary" Fiero. This particular car is owned and driven by Paul Hosler. Before we continue, have a look at the rest of the pictures. If you know anything about the IMSA Fiero, you will notice immediately that Paul's car isn't original. He modified it to take a V8 with a Porsche transaxle. Yep, that means that the drivetrain now sits longitudinally in the car. Notice also how clean his car is. And his garage. My garage has never been this tidy.

I first got in touch with Paul as I was trying to find information on the IMSA chassis. I figured it would make an interesting CAD project and Paul sent me a scan of the drawing for the rear upright just to let me have a taste of what I was getting myself into. Being me, I couldn't resist to model the upright and send Paul an IGES. Apparently, this was appreciated as Paul made me an offer I just couldn't resist: he would send me drawing scans and measurements and I would model the original chassis and his modified version. The plan is to compare the torsional stiffness (between the original and Pauls version) and see if any improvements can be had.
Now, this is where I would post pictures of the CAD model, but I'll save those for the next post.

Big Brake update

I have had two adapters water-cut to their basic shape in 20mm steel (actually, they measure slightly thinner than 20mm) and have drilled and milled the first one according to my CAD model and drawing. It does seem to be a perfect fit. Now I have to do the same for the other one and then give them some sort of surface treatment. Galvanizing and painting perhaps?

For the record, the material for the adapters is steel with a yield stress rating of 355MPa. As far as I can remember, my last analysis gave a max von mises stress of 200MPa. And yes, steel is denser than aluminium and my adapters are heavier because of it. But there are other considerations as well. Besides, why worry about the small extra weight from the adapters being made from steel instead of aluminium when I'm adding huge cast-iron rotors anyway.

Vette caliper modelled

A semi-accurate (hand measurements) model of the vette caliper has been added to the assembly:




before ordering the rotors, which I nearly did, I realized it might be a good idea to see what the guys over at rejsa.nu would say. After all, they have some experience with these sort of things.

Corvette brake upgrade adapter - second revision

As you can see, I did a bit of work on the design of the C4 brake upgrade adapter. The model of the rotor was modelled from measurements given to me by Art at corvette central. I guess I should be placing an order for those rotors soon now that I finally have the money.



Here we have a slight quirk. Although I consider this design better than the original design mainly because it makes use of the original threads (the original design, which can be found over at pennocks fiero forum or bubbajoe's website, required the threads in the upright to be removed and the adapter threaded to take the M12 bolts) it's not without quirks. I think I will make use of longer bolts and add nuts behind the upright to secure them. Probably not needed, but better safe than sorry.

The virtual guitarbuilding project

If you have visited iguitarer.blogspot.com you might have seen this project. As I've said, this blog is for showcasing my projects and it would be a shame not to mention this considering the time spent on it. These pictures are of the assembly, which is of course the easy part, but they give a good overview of it all.









For more information on this, head over to iguitarer.blogspot.com and search for "virtual guitarbuilding"

Caliper bracket analysis

As you can see, I put my bracket through the FE analysis of my CAD package. The intention was to get loads to use with the adapter.

So what did I put into the analysis then? Well, for restraints I used the normal type of constraint. The two faces which are in contact with the adapter were locked from sliding along the adapter. On the opposite side, the edge of the hole was locked from moving perpendicularly to the bracket. I realize now that this version is not entirely true as the bolt is threaded into this hole in the bracket. well, I'll see some other day if it makes a difference.

The load was applied where the pad backings make contact with the bracket.

The load was determined by a few calculations. First, I aimed for a stopping distance of 30m (which is unlikely, but something to aim for) which means an average deceleration of 1.3G. I then calculated the balance at this rate of deceleration which gave me the normal forces acting on the wheel. The formula then was:

my x N x D/2 = F x d/2

where "my" is the friction coefficient between tire and ground (1.3 to reach the deceleration rate I aimed at), "N" is the normal force acting on each wheel, "D" is the wheel diameter and "d" is the effective diameter of the rotor. "F" then is the force acting on the bracket.

I calculated F to be about 13.2kN. However, for the analysis I set F to 15kN.

And the end result is the picture at the top.

Putting a stop to it

Behold, my current car project. "What is this?" you ask. Well, this is my take on the Corvette C4 12" brake upgrade for the Pontiac Fiero. This swap was described in detail over at the fiero-forum (fiero.nl) and was an obvious choice. But there were things I did not particularly like about that swap. The major thing was that it used M12 bolts in place of the original fiero (M12 fine) and Vette (M14) bolts, requiring the threads in the fiero spindle to be drilled out. So what you see here is my first attempt at improving the adaper.



So today I got some information from Gus at corvette central regarding the C4 brake rotor dimensions and the difference between the 84-87 and 88-96 base rotor (11,5 vs 12 inch diameter and a difference in offset). Once I had this, I included it in my assembly:





No, that is not an optical illusion. I have a design flaw, causing rotor to bracket interference. Well, that will be an easy fix. I think. I also need to add a wart on the adapter to act as a steering stop.