MAN Diesel engine from 1907 "DM 2 * 100"

[MuellerNick]

Casting #14. Guide for the plunger of the fuel pump.

Nothing exciting. I have cast that part in sand before, but it got too blurry.
Disregarding the fact that I had to remake the core box (ruined it in the last 5 minutes of milling), all went well:


Core box. Sorry, no core, used them up before taking a picture.




4 on a tree. Makes machining easier.
Also made these with vacuum assisted casting (currently my favorite process).


Nick

 

[MuellerNick]

I have a little more to show[tm]!

Had to re-make the core box for casting #8. I misplaced the alignment cones, so they were mis-alignment cones and the casting was offset by 0.8 mm.


#8, version 2. As I had issues with shrinkage, I reconfigured the runner/gate part. Reduced volume of the gate and increased the surface of it. That pushed the shrinkage where it doesn't hurt.

Casting #15. Oil pump.
This is, I thinkt the smalles part. It has two separation planes, and milling them at that dimension requires some planing ahead, because you have to consider the width of the milling bit and still have to get it right.


As cast. The red line shows where the part will be cut off. These two "antennas" (1.2 * 1.5 mm) are artifacts of shifted separation plane.
The oval detail is 1.5 mm wide.


An other look at it, at the two separation planes. Here you see why they are needed.


Core box and core. Arrows point to the shifted plane.


Casting #16. An other different arm for the generator. This one holds the timing ring.


As cast, all features still in place.


Other view. Shrinkage is where I wanted it to be.


Core box and core.

The parts of todays postings were made this week.

Next, I'll post some pictures that show "assembled" subgroups.

Edit:
All parts of this posting were vacuum assisted castings.


Nick

 

[kvom]

The casting stuff is fascinating, but I can't wait to see how you finish machine these parts. They don't look too easy.

 

[MuellerNick]

but I can't wait to see how you finish machine these parts.

You and I have to wait until I have all castings.
At night, I dream of machining setups. Some end in nightmares.

I will make soft jaws for all milling ops. That are jaws for the vise that have milled pockets for the parts to go in. Thus, I do have a precise reference (jaws are indexed).
I used that, and it works great. You can even clamp painted parts and mill them, close the vise with full power (40kN) without hurting the paint (as long it is a decent paint; I'll use 2K PU industrial paint).


Nick

 

[MuellerNick]

Thought I could make a sub-assembly.
So I took the biggest file and some CA-glue.


For your convenience, I have written the casting # and the page of this thread close to the part. In case you want to refresh your memories.


I tried to find a picture that shows all parts involved. Won't be possible for all following sub-assemblies. But this time it worked.



Roughly the same perspective like the original's photo.

You have to bear in mind, that I didn't remove all finishing allowances (most of the time 2mm). So parts look distorted, offset or a bit wrong.

Edit:
Dammit! I just saw that I placed the fuel's pump nipple at the wrong place. Well, this will be the only mishap I guess.


Nick

 

[canadianhorsepower]

this kind of work excedd what my medication is allowing me to see
awsome work

 

[skyline1]

Hi Nick

This is coming along at an amazing speed, how do you make all these patterns and core boxes so quickly ? It would take me simply ages to make all those. Do you CNC machine them or 3D print them or something. Even a professional pattern shop would be hard pressed to keep up this kind of output by "old school" patternmaking.

I like your magnesium reheat technique and your vacuum assisted casting setup is super.

A commonly used practice in foundries to help with shrinkage is to use extended sprues or feeders which give the metal more "head" and a body of metal to "Draw" from as it solidifies. I am probably "teaching my granny to suck eggs" on this one so I apologise if this is something you already know.

Keep up the good work at this rate you will have a finished engine before any of us can blink.

Regards Mark

 

[MuellerNick]

This is coming along at an amazing speed, how do you make all these patterns and core boxes so quickly ?

Some work is realtime, other is old work.
Those vacuum assisted castings are realtime (but not showing them every day). The ones shown take about a day from start to finish. Milling time is about 3 … 4 hours. But before, I have to generate the G-code.
After the core box exists, I have to make the cores. Takes another 2 hours. Because when I'm at making cores, I also do make others.
While making cores, I melt small crucibles for pouring the moulds.

Other patterns (for sand casting) take machining time from 5 hours to 40 hours. Finishing them an hour to a few days. I have to make a test pour to see where there are faults. Some patterns needed about 4 tries.

A commonly used practice in foundries to help with shrinkage is to use extended sprues or feeders …

Those vacuum assisted castings work a bit different. The melt cools down quite quickly. And it has two gates. The first one going into the mould is just 3 mm in diameter. If I would make that bigger, the finish suffers because the melt rushes in too quickly. So the cavity is cut off from the pouring hole after a second or even less. Thus shrinkage will always be inside of the mould. I just have to get it into the runner/feeding system. But I'm getting good at that.


Nick

 

[MuellerNick]

Valve-related castings ...

Casting #17. Coolant in/out for the blast air injector body:

This one is a vacuum assisted casting.


But they work the same when I embed the core sand moulds in oil bound sand.


Core boxes and cores. Not the most economic design, as this takes two different cores. I just could have drilled the missing gate. Well, I learn along the way ...


Casting #18. Valve body for inlet and/or exhaust.


As cast. Gets cut off at the red line.


An other view. The mould is made of 3 parts.


Core box and core for part 1.


Core boxes for part 2 & 3.
Also not one of my most clever designs. I should have added alignment cones. But I learned that later.


Cores of part 2 & 3, roughly held together to show how they mate.


Part 1 & 2 held together, a look into the cavity. The outside of that mould is a bit odd-shaped. But who cares, only the inner values count.


And that's how it looks like, if the mould bursts with vacuum-assisted casting. This happened inside the vacuum-chamber, so nobody harmed.


Nick

 

[gus]

Hi Maestro.

Your thread very educational. Was Nick a professional foundryman? Your patterns and sand molds very well done.
The best Ferrous Foundries we have in USA were run by German Ethnic Americans. Help audit Great Lakes Iron/Steel Foundry in Michigan. They still operated Cupola Furnaces. Quality and C.I. composition very well controlled.They use US equivalent "Toyoda"
Auto Squeeze Mold Machine.
Gus now planning to make a mini Gas fired Aluminium Furnace.Now also looking for Oil Sands.Hope I won't end up burning my apartment.

 

[MuellerNick]

Oil sand in your appartment? You will regret that!
Casting in oil sand in your appartment? You will regret that!

I never learned that trade, I started casting in last year's spring.

Nick

 

[gus]

Oil sand in your appartment? You will regret that!
Casting in oil sand in your appartment? You will regret that!

I never learned that trade, I started casting in last year's spring.

Nick

Hi Nick.

Please tell me more about oil sand. My aluminium casting limited to gravity cast with cast iron molds for qty of 100---200. Did watch my uncles using green sand.

May go into clay sand. Or perhaps Clay/sand/cement.

Your expert advice appreciated.

 

[MuellerNick]

So I got some news ...

I wanted to cast a new part today (Compressor), but then found that I made one of the casting's core sand boxes wrong. 3 hours milling time in the bin. Even before casting a single piece. :fan:

So here we go with new and old parts:

Casting #19. Injector main body.


As cast. Core sand mold, embedded in sand works best with this.


Core box. I have no cores left, all already assembled.


But I quickly machined one for you. Just drilled through and faced off at the bottom.

Assembly (with CA) of the cylinder head.

I had to leave out some parts, because I'd have to machine them completely to make them fit. It is quite crammed on the cylinder head. So only the exhaust valve body, the injector, coolant connection for injector, rocker arm for injector and the two rocker arm shaft posts.

And a new casting. Casting #19. Coolant passages for the compressor to come.


6 in one go. The footprint of one of them is 5 * 10 mm, about 4 mm high. They will be cut off at the black line. The two rearmost ones failed (sand broke).


Core box and core. This one is the most pickiest to get out of the core box, as you can see by the traces from hammering.

Nick

 

[gus]

I admire your craftsmanship and dedication to detail. Are you retired and now pursuing this hobby full time.
The foundry patterns would have taken a lot of time to make.
Wish I have the time too but I have too many other hobbies. Fishing takes over the weekend.

You are the Maestro!!!!!!

 

[MuellerNick]

While the putty on the next pattern hardens …

… I wanted to come back to #19 to explain a trick:

If you see such a casting, you will think it is impossibly to do. Maybe with spin casting in pewter. But not in aluminium with sand.

Well, you see that it works. And I'd bet that this part can be cast in oil bound sand too.
The very simple trick is to have much more material in the back of the desired part. This helps to get enough heat (or: prevents cold material) into the desired cavity.
If I'd have to cast this wee part in oil sand, this would be the design-procedure for the pattern:
Start with the finished shape.
Add about 2 mm to the back (the face that faces my finger). Just enough room for a 1 mm slitting saw.
Repeat the detail for maybe 4 times (you'll have rejects, and you need more of those).
Add a massive block of material behind it. That block can well be 20 mm thick, and, looking from the top, be 5…10 mm wider all around. This eases pulling the pattern from the sand. That block acts as a guide/alignment feature when pulling the pattern and makes carving a runner to the cavity an easy thing.

You also should think about work holding when designing a pattern. It doesn't help that much if you have a tiny thing that needs further machining, but there is no way to clamp it in the vice. The waste of material won't kill you. You can re-melt it.

What is the difference between a baker-man and a foundry-man?
The baker-man has to eat his rejects, the foundry-man remelts them.

I even came up with a procedure for to make tiny fake-bolts. That part needs 4, one in each corner.


Nick

 

[MuellerNick]

So I made a test-cast of the compressor today.
Even if it was a failure, I count it as success.

I have an issue with shrinkage, but I wanted to give it a try.
I also wanted to see, wether a steel core would come out easily (without taper). It doesn't. I expected that too.
The cure for both is making a core out of sand, the way I initially thought I should do.
But you don't learn if you don't try.

The pattern needs refinements in two places, but that is no mishap, it happens often (at least to me).


So here we go …
Casting #20. Compressor.
This is a two staged compressor. Diesel initially had planned a one stage compressor in his previous engines. But they simply worked as Diesel engines, as they had a CR of about 1:70. The oil that was unavoidable for lubrication ignited from time to time.


As cast.


An "impossible" feature. The three cylinders sticking out. The middle one is no problem, as it is in the splitting plane. But the upper and lower ones prevent a simple split pattern to go out from the mould. Without breaking it, that is.


Lower end of the casting. It will be cut off at the black line, machined and both parts bolted back together. So that is two castings as one. Yes, I added enough machining allowance.


The top with the compressor's head that I made wrong and the wee coolant passage. All roughly cut to size, so no perfect fit.


The pattern with the tool to pull the inserts. I'll come back to this in a moment …


Core box and core.



I needed a way to precisely index onto the lose inserts' axis. So I added a pointy cone that was milled 3D (and 3-axis). Tilted the head and matched the center drill's point with the pattern's cone. Set all to zero and retracted with G0 X-50 Z-50, put in a drill and drilled with G1 X0 Z0.


Pattern in the sand. Before pulling it, I first have to un-tighten a grub screw and pull out the insert. Then the pattern moves out.


Mould with core and steel core in place. Here you see, why I had shrinkage. There simply is too much material at the right side and there is no way to add a feeder. So I need a core the way it has to be.

I talked about embedding core sand moulds in oil sand. That's the way it looks. On the cope, I just carve rough runners that cover the holes in the core sand moulds and that's it.



Enjoy!
Nick

 

[canadianhorsepower]

did someone check your DNA to make sure your a NORMAL human being
awesome work

 

[MuellerNick]

did someone check your DNA

Last time they checked, they said that my gears did get some backlash, but my compression ratio is very good.

I forgot, I do have a video about the castings #4 and #7, both on page 3 of that thread.
SmallCastings

But I changed the process for #7 (the lid) to vacuum assisted casting. The core sand mould is still the same.


Nick

 

[Diversion900]

Magic to watch it all come together, those castings are a work of art in their own right.
I am following along, fascinated.
Keep up the good work....

Cheers, Neil

 

[MuellerNick]

So I had to make some repairs.

First of all, my heating. Got a bit cold.

Then, I repaired/fixed three castings:

Casting #10 of page 5, where I misplaced a feature.

Taking the chance to add a feature. The tiny tongue (3 * 1.5 mm).


Other view. The core box needs a bit of finishing up.


And I changed it to core sand mould embedded in sand (vacuum doesn't work that well there).

The compressor head. Casting #21. Have already shown that one, but it was wrong, so got no number yet.

Casting.


Core box and core. Cast embedded in sand.


And a better core for the compressor.

Core box and two cores glued together. Yes, it is long and slim. No, it does not break.


Cores in the mold. This time, I placed the upper core the right way round.
Still have a slight shrinkage in the area at the red circle, so I'll change the upper core. Not that tragic.


Nick