Learn me reverse engineering, specifically machined and cast parts.

I've been teaching myself casting recently, I've set myself a goal as a sort of Journeyman's Project.  I want to cast my own cylinder head for either my car or truck.  

To start with, I was going to get a feel for how the stock head looks inside, to do this I was thinking about

1. getting a junked head
2. setting it up on a mill
3. cutting off about a tenth of an inch per pass (like this https://i.imgur.com/cN9WjmPh.png )
4. recording the face (either through a picture or using it like a stamp)
5. CAD it up and get a model of a stock head
6. start modifying
7. 3D print, cast, be awesome

 Anyone have a better idea of how to do it?  

No idea but sounds cool. 

You will need to be real good at making cores ...

In reply to einy :

I'm not so concerned with the actual casting portion of it, I've got good help with setup and such.  Its the getting to having something to cast that I'm concerned with.

That’s cool ... maybe pick a simple head to start with like an older GM 4.3 pushrod V6 or similar.  Those are about as simple as they get, and don’t even need inserts for seats or guides.

I can offer you a few things since reverse engineering is a significant portion of my day job

• First off, what are your goals?  Are you looking to make stock replacement heads for something unique or are you looking to make a hotrod part?  
• What ancillaries that bolt to the current head are you looking to keep?  Sensors, threaded bosses for accessories, stock intake or exhaust manifolds?  Cams?  Valve cover(s)?
• Honestly the easiest way to find CAD data if its a common item is to download it from a repository.  Search around, there are either some free or pay models available that would help you.  I downloaded a complete GM LSx motor a while ago just to play with, along with a complete Toyota Land Cruiser exterior model.  Cool stuff is out there made by people way more talented than me.
• If I was starting from scratch and I actually wanted it to work, I would pay the $300 or so to have it 3D scanned on the exterior to get all my bolt holes in the correct location.  Then I would bandsaw cut it vertically through the port centerlines and start after it with a set of rulers and radius gages, or have it rescanned again.  
• Once you get a model then the real fun begins.  First, you have to add shrink to it.  Then you have to gate and riser it.  Aluminum shrinks around 1/4" per foot but you can get more accurate numbers online I am sure.
• Once you get a model, get your shrink done, add draft, figure out your parting line, get your gating, risering, and core prints done, then you have to make all the core boxes for the ports and also the main pattern.
• Shell cores will give you the accuracy you need without having to CNC the ports afterward but you are going to need to have those core boxes cut from iron and made by someone with a shell core machine.  Most large cities do have a core shop, or get in touch with Supreme Cores in Milwaukee.
• You can also make ramup cores with airset sand and glue them together but you wont get the accuracy you need to not have to clean the ports up with CNC afterward.
• Now you get to make a pattern and from that, a mold.
• Alloy-up and pour your aluminum.  
• Get it X-Rayed and make sure its sound
• Get a layout on it (again, due to complex geometry suggest scanning it and having the part checked against the model)
• If its all good, then you get to heat treat it, machine it, get the valve guides and seats pressed in, valve seals, etc.  
• Put it on and run it!

Best of luck, its not an easy project but can be very rewarding.  Any other questions feel free to ask!

Casting is a rabbit hole.  Enjoy.

I'd start with a Briggs and Stratton flathead from the 60's and work my way up from there.

This guy had a fun project that might be relevant:  https://www.youtube.com/watch?v=jVD-P-77B24

just as an idea , if you do a small part , use a  desktop paper scanner , 

Then you can do your idea of cutting  100 thou off , scan again , repeat , repeat......

this will give you a lot of data points , 

others have made  foam copies of the cores  for the intake and exhaust,

if this is for learning , I would do a head that you can get a few of so you can check your work , 

and if you do a smaller part you can 3D print it , nit easy to do with an Chevy head

Sounds Fun as a learning tool , 

I think your basic plan is sound but agree that maybe you don't want to cut such fine layers at first. You might have a lot of edge burrs that could impede your productivity or accuracy. Maybe slice like bread at 1/2" layers to "explore" and then decide what kind of layering you want to do on a 2nd head. And I agree with 93gsxturbo that you're going to need the bolt hole locations 3d'd somehow before you start slicing or the ones that you slice through axially you'll never know where their centerlines were.

Do a dual port 1600 vw head , 

loads of them around ,  nice size to work with  ,  will fit on the scanner ,  8 head bolt holes all the way thru so you can line them up when you have the slices ,

Plus you get the fun of figuring out how to do cooling fins !

If you have a source for 3D metal printing, it may be expensive, but fun to make a head from.  

There is also tech to 3D print casting cores.  I've heard of it, but for what I really do, it seems about 5 years late for what I see them being applied to.  Still, something to look into.

here is an interesting youtube video that casts a cylinder head using 3d printing

https://www.youtube.com/watch?v=pFfOxNulazg

93gsxturbo said:

I can offer you a few things since reverse engineering is a significant portion of my day job

• First off, what are your goals?  Are you looking to make stock replacement heads for something unique or are you looking to make a hotrod part?  
• What ancillaries that bolt to the current head are you looking to keep?  Sensors, threaded bosses for accessories, stock intake or exhaust manifolds?  Cams?  Valve cover(s)?
• Honestly the easiest way to find CAD data if its a common item is to download it from a repository.  Search around, there are either some free or pay models available that would help you.  I downloaded a complete GM LSx motor a while ago just to play with, along with a complete Toyota Land Cruiser exterior model.  Cool stuff is out there made by people way more talented than me.
• If I was starting from scratch and I actually wanted it to work, I would pay the $300 or so to have it 3D scanned on the exterior to get all my bolt holes in the correct location.  Then I would bandsaw cut it vertically through the port centerlines and start after it with a set of rulers and radius gages, or have it rescanned again.  
• Once you get a model then the real fun begins.  First, you have to add shrink to it.  Then you have to gate and riser it.  Aluminum shrinks around 1/4" per foot but you can get more accurate numbers online I am sure.
• Once you get a model, get your shrink done, add draft, figure out your parting line, get your gating, risering, and core prints done, then you have to make all the core boxes for the ports and also the main pattern.
• Shell cores will give you the accuracy you need without having to CNC the ports afterward but you are going to need to have those core boxes cut from iron and made by someone with a shell core machine.  Most large cities do have a core shop, or get in touch with Supreme Cores in Milwaukee.
• You can also make ramup cores with airset sand and glue them together but you wont get the accuracy you need to not have to clean the ports up with CNC afterward.
• Now you get to make a pattern and from that, a mold.
• Alloy-up and pour your aluminum.  
• Get it X-Rayed and make sure its sound
• Get a layout on it (again, due to complex geometry suggest scanning it and having the part checked against the model)
• If its all good, then you get to heat treat it, machine it, get the valve guides and seats pressed in, valve seals, etc.  
• Put it on and run it!

 

Best of luck, its not an easy project but can be very rewarding.  Any other questions feel free to ask!

Goal is definitely to improve over stock but it would almost be impossible to not do so, both of them are asthmatic as hell.  The truck I could cut up a pair of LSx heads (that's another thread) and make an I6 head out of them in order to solve the breathing problem, but the Mustang doesn't have much of a solution (unless I want to spend $2000 for a head and all the trimmings)

Ancillaries are up in the air.  Only one of them has a sensor in the head from what I remember and that one doesn't even have an intake pattern to bolt on.  It will depend on what I find and what I can upgrade.

Part of the reason I want to do this is because there is nothing (I can find at least) on these heads or engines online.  The bolt holes I was figuring I could get from the head gaskets and intake/exhaust gaskets.  Valve cover would be a bit of fun though.

I've wondered about 3D scanning.  I've got a couple contacts at a couple places nearby, I'll reach out and talk to them about it.

In reply to californiamilleghia :

That was damn sexy.

alfadriver said:

If you have a source for 3D metal printing, it may be expensive, but fun to make a head from.  

There is also tech to 3D print casting cores.  I've heard of it, but for what I really do, it seems about 5 years late for what I see them being applied to.  Still, something to look into.

The latter is kind of what I'm thinking of.  For a production run it would be ruinous.  For my purposes...

Seems like a fun project.  I agree with your approach for the most part.  Get all the hard points.  From whatever measurement method.  

For the guts, I would cut into just a single cylinder like you suggested, remove the burrs, put it on the scanner.  Put something like a coin on the same scanned image that you can measure with calipers so you can scale the image accurately.

For the next cylinder do the same thing in the other direction (along the centerline of a port)

Bring those images into cad and arrange them based on their actual heights/depths.  Trace and loft/extrude/etc. as needed.

3d print and test fit everything.

Cast and enjoy.

In reply to ProDarwin :

I'm waffling on the scanning.  It would probably be very easy to mess up the setup of the cutting head such that I didn't get a good cut.  Maybe with the right jig...

I was actually going to get somewhat elaborate with the picture portion of it.  Rail system to keep the camera at the same orientation and position, something like a photo scale ruler (the right angle rulers you see when forensics teams take pictures of evidence,) a fixed lens on the camera, and a standoff to keep the camera the same distance away every time.

No sense doing it if I'm going to have to worry about parallax, angle of the camera, and other factors.

How much experience do you have machining?  Noone with experience says a "tenth of an inch"...

My thoughts are you need to spend sometime in a shop.

In reply to Fueled by Caffeine :

Machined the jigs for every single suspension mounting point on three FSAE cars, one of which won 6th at Lincoln.  Several internships in machining and fabrication companies before I landed my actual engineering job.  I'd say I have enough experience to know what I'm doing when it comes to this.

I'd also say that a mild learning disability sometimes has an effect on the words or phrases I choose.  I say what makes sense to me and worry about clarification afterwards if it doesn't make sense to others.

What would you say, "a hundred thou"? That doesn't make sense to me, so I dont use it.

I'm not a machinist so I don't speak like a machinist.  Doesn't mean I'm incompetent. 

In reply to Mr_Asa :

ok.. then do it to it.. but I just dont want to get someone hurt by watching a few youtube vids thinking they can throw chips...  I had to pull a guy out of a lathe once.. it was bloody. this is not stuff to be messed with lightly.

Mr_Asa said:

In reply to Fueled by Caffeine :

What would you say, "a hundred thou"? That doesn't make sense to me, so I dont use it.

I'm not a machinist so I don't speak like a machinist.  Doesn't mean I'm incompetent. 

X2

In reply to californiamilleghia :

That was fun.

For your modifications, perhaps find an additional junk head to cut up that's closer in design to what you'd like to replicate, and compare that to your findings from cutting up the stock head.  Especially if changing from iron to aluminum and/or wanting to significantly increase performance.

Other than that, I look forward to (stealing ideas from) your project/build thread! 

Mr_Asa said:

In reply to Fueled by Caffeine :

Machined the jigs for every single suspension mounting point on three FSAE cars, one of which won 6th at Lincoln.  Several internships in machining and fabrication companies before I landed my actual engineering job.  I'd say I have enough experience to know what I'm doing when it comes to this.

I'd also say that a mild learning disability sometimes has an effect on the words or phrases I choose.  I say what makes sense to me and worry about clarification afterwards if it doesn't make sense to others.

What would you say, "a hundred thou"? That doesn't make sense to me, so I dont use it.

I'm not a machinist so I don't speak like a machinist.  Doesn't mean I'm incompetent. 

Wish people would toss this kind of context out when starting the thread. The sandbagging does not help focus replies.

Up till I read this I was thinking this was a fool's mission ( nothing against those) Now its a case of "You probably know the answers already so why ask?"

In your situation I would go straight to trying to cast anything just to get my feet wet with the chosen technology, tools and materials, and scale from there. I have zero idea of what tools, digital or otherwise, you have at your disposal to work on this.

There is someone on this forum who is trying to cast an intake manifold, and while fascinating to watch each iteration, I don't know that success has been declared yet. Perhaps they will chime in.

Pete