GTgears

You're talking about the **** "RSR" box? Can't you put a 991 box in the car?

cobalt

Not sure what aluminum alloy standard you are working to I might suggest a copy of the old QQ-A-601 or ASTM B-26. If you don't have a copy I can scan them and send them to you. This might answer some questions although there are many different specifications these are the primary ones used for Aircraft applications.


I'm partial to Brinell for testing aluminum and magnesium however Rockwell is perfectly acceptable. Some companies don't allow cross referencing but it isn't a concern. My old Osborn hardness tester was adjustable up to a 5000kg load and designed to test all types of materials. This was WWII vintage stuff built to last 10 lifetimes. I used a 10mm indenter ball which is specified for MIL work and gives most accurate results but what you have will do fine. Just remember to always inspect the indent at 0 and 90 degrees and average the two and at least half inch of thickness, but the requirement might be less for a 100kg load. Sounds like you know that already. Also you will see different readings around any in-gates or risers.

Hard to say what is going on with the magnesium case. Could be anything from quality, material, age, how it was treated over the years etc. I might suspect that the 1/16" ball size might be a bit small for mag. It makes a very small indent and magnesium being softer in general so using a 10mm indent ball would be able to give you more consistent results. Mag will see numbers on average to 70 BHN vs aluminum which will see closer to 85-90 BHN on average. I achieved very strong mechanicals from my mag parts but I oversaw or processed all of my parts personally as that is a process I find many don't get correct with magnesium. (not to mention "fire") The nature of the material will have less yield strength than aluminum but can have equally strong Ultimate Tensile seeing consistent numbers from 38k to 42k PSI.

Normally I would see consistent hardness numbers throughout my parts and although Mag will vary more than aluminum depending on the size and shape of the parts you might see some thicker sections show to be softer than the thinner sections. This is in the case of mag castings that might have cross sections 4 to 6" thick vs other sections that are .5" or so. Aluminum is different due to the rapid water quench used.

I am sorry to say I am not well versed in indent loads at 100kg or how difficult it is to read a 1/16" ball indent. Reading 10mm indents under high magnification I'm assuming is much easier albeit takes practice to get right. I believe that is why most MIL specs specify a 10/500/15 BHW.

If you want to take this off line I would be happy to try to answer any questions you might have. I make no promises as doing anything from afar is always harder than in person.


Thanks for rubbing it in. It is turbo weather already and the leaves are changing very quickly. Another week or two of swimming then it is time to close everything up for winter and start thinking about putting the cars to sleep for winter.

Was about to start making mag parts for 962's and 935's when health and other factors changed my direction. But I learned things happen for reasons that aren't always clear while you're living through them.

Catorce

Cobalt,

PM'ing you now with my info. Good stuff!

Edgy01

Wonderful! With CAD-CAM and CNC machining at your disposal you can produce something far better than anything Porsche envisaged back in the later 1980s! Keep in mind that you have a lot of detail to deal with in that case, so don't expect it to be perfect. That's something you can deal with in the future. I would think the most important elements are dimensional compatibility with other component producers, etc. At some point you will want to improve it again, and again.

I really like the issue of the strengthened aluminum. Aluminum has many favorable qualities, but your method appears spot on. I recall when Porsche had been so obsessed with weight that they spent years on magnesium cases, only to finally switch back to aluminum for the 911SC cars in MY78. Whats a few pounds compared to reliability!

Keep up the great work! I'm happy to hear that there are smart guys out there thinking of this stuff!

Another technology to consider--read up on it. It may be useful in places. "Super plastic forming." We've done it with titanium, however.

misterwaterfall

Makes 2 of us. Super interesting discussion though from what I can comprehend.

Catorce

Edgy,

Thanks for the kind words. My tack is a little different, however, and it is imperative that the first cases are functionally perfect.

Being the first one to market, there is a high bar set right from the get go. So far, our castings turned out perfect....or we kept trying until they were. Machining - same thing. it is my goal that no customer will ever receive anything defective or short of the mark, and that is why this project is taking so damned long.

Future improvements are a given - we will find better ways to do things. But case number one will be 100% functionally perfect, since it is the case I am putting in my 73RS.

It WILL be perfect. That is the mission.

Falk930

Catorce,
my biggest respect for your work!

Honestly I don't know a lot about casting, but maybe I missed something
or its not important at all:
You measured one, I guess used, 964 case very precise.
Or better you digitalized it and verifyed that very precise.
From that you build your 3d model and production route.
Evey step in your production need tolerances.
Small as possible/ big as payable.
But you don't know, where your reference case was at upper or lower
production tolerance.
We call that chain-tolerances, had one case in a far more simple
situation.
We measured a sample, added production tolerance
and the part didn't fit.
Reason: Sample was produced at the upper tolerance.

Thanks
Falk

Catorce

Falk,

100% correct. There is a much wide range in tolerances for the Porsche cases than there is for ours, since they produced so many more cases of a period of many years.

So what we do to eliminate those kinds of errors is pretty simple. We scan the whole case, and before I go on, consider that about 80% of the case consists of features that are NON CRITICAL. The placement of the case perimeter bolts is not critical as long as both sides match. The angle of the curvature of the thermostat housing is not critical as long as the thermostat fits inside.

So basically, you have a SLEW of non critical numbers that your scan picks up, and the scan is pretty accurate.

Now the good part - the critical numbers are published in the spec books. Bore diameters, spigot diameters, journals, layshaft to crank centerline etc.....all in the book. So with these types of numbers, we plug in the actual hard number into our model.

Therefore, the end product is going to be pretty darned accurate. Of course it will get lots of testing as well, but we're confident we have not accumulated any tolerance buildups.

TexasPorschelover

When can I buy one?

Catorce

Texas,

That's the million dollar question but I hope in January at the current rate.

Catorce

It looks like we won't be able to meet the schedule posted in the last post. Making the casting is one thing, but the machining is proving to be a major area of focus right now.

In some respects the casting is the easy part, although thus far nothing in this journey has been easy. The reason why I say the casting was the easy part is that there is a lot of extra meat on the casting, which of course needs to be machined off. This extra meat makes the casting somewhat more forgiving to make.



The machining, by contrast, has to be ultra precise. To that end, we have rejected many of the machining methods used to machine the case, and even rejected many of the machines themselves. Originally, we were going to machine the cases with a combination of vertical CNC mills and CNC lathes. That idea has proved itself to require too many setups, which affect the repeatability and the accuracy.


We have now realized that the machines we had access to won't be appropriate for mass production, and so have decided to use a combination of large horizontal CNC mills and a 5 axis Vertical CNC mill.


These machines can machine our part with fewer setups, and less movement of the part than the machines we had access to, which will of course result in a more accurate part, and more importantly, a more consistent part.



Thanks for your patience as we endlessly fiddle with this, and rest easy in the thought that once we nail the routine down, the actual production of the cases will be elementary....

Luckyboy

Subscribed!

Clive Jameson

I'd love to hear how the process is coming on this project. I hope you're clearing the hurdles as they inevitably come up.

Rod_Ez

I've been following this thread since day one. I can't wait to see the end result.

cobalt

Here is an interesting read for those following. Although I love the technological advancements made they make it sound like the entire science is something new. I was making castings like this thirty years ago using hand vs virtual processes with the same level of quality. I do miss the challenges.

​​​​​​http://www.racecar-engineering.com/a...locks/#Pouring