MAN Diesel engine from 1907 "DM 2 * 100"

[AussieJimG]

It just keeps getting more amazing

Jim

 

[MuellerNick]

Oh, here is a very short video of a smaller one, but also from the "DM"-series:
[ame="http://www.youtube.com/watch?v=VTppEG59xcs&feature=endscreen"]YouTube[/ame]
Maybe the 20 hp-version. Hard to say without an overview. Certainly not the 8 hp version, as that one had the A-frame and the base plate cast as one piece.

You will recognize the fuel-pump and the little feature I added. Also to be seen is a part of the compressor (at the left) with the coolant-passage (the smallest casting so far). It is just above the spring sticking out the fuel pump. Stop the video at 0:33 to see better.

Edit:
And here is the smallest version, the DM8.
This engine can be seen in Augsburg, where MAN has his museum and archive.
[ame="http://www.youtube.com/watch?v=ykKooloIafg&feature=related"]YouTube[/ame]


Nick

 

[MuellerNick]

Re-reading the whole thread, there are some questions left unanswered:

Bore/stroke is 450/680 mm. That makes for 108 liters displacement. At an scale of 1:10 that will be 45 * 68 mm with 108 cm^3.
The plan is to make a true running diesel.
But first, I'll make a static display model with two cylinders as shown. All parts moving, but not running on its own. Next step will be the one-cylinder version. On that one, I'll try to get it running. I do not intend to place the fuel pump etc. where the original is, because that would not work, or in other words, I'd have to leave the scale appearance. Most important part is the scale. Functionality will be hidden in the model's base. So all the injector things, lubrication etc. will be in the base.

3D-printing for patterns:
All 3D-prints I have seen are not worth considering as patterns. Way to rough and staircases in slightly off-horizontal areas. They would need more rework than milled patterns and cost more money.

I'm using block material from Renshape. One (the beige one) is with 700 g/l density for the core boxes. The other one (green) is for matchboards (funny number hard to remember). The beige one is more sensible than the green one, but cheaper. Until now, I have spent 600 € on block material (and will have to order new one).

If I have left out to answer question related to that model, just ask again or PN me as a reminder.


Nick

 

[MuellerNick]

The reason why I started vacuum assisted was this casting:

Casting #22. Protective cover for compressor linkage.
Since I started experimenting with the vacuum assisted casting, that thing never worked. First, because of wrong parameters. But I found them two weeks ago. And second, because for an still unexplainable reason, the inner core was too big. So I took that one, measured, compared it with the CAD. Too big. Verified it against the G-code, too big also.
So today, I milled it again …


As cast, with the gate still in place.


A old reject, with the gate sawn off. The problem of that casting is the thin shell (2 mm) and the fact, that the melt has a long way to reach the end of the cavity.


3 core boxes. The lower one is for the core. A core out of core sand to go into a mould out of core sand.


The cores. The core I'm holding in my hand is the core. The half cones match into the core moulds and align that core within.


One half of the core mould with the core sitting at its place. The slanted face at the lower left, I have to manually sand down. Or the core box would have been gotten more complicated. Sanding down is quick, I just rub the part on the table, the core provides its own sand for sanding.


And today, while having a walk to lunch, I made an invention.
DTC. Short for DeskTopCasting.
You already know, that I do embed core sand molds in oil sand and cast them there. This is for castings that need more feeding like the compressor head. They don't work with vacuum assisted casting (unless I change the mould and make it much bigger).

The idea was to cast a pot for the melt that is glued on the mould and provide enough room for hot metal to feed the casting while cooling.

Well that's such a pot glued to a mould.


Two tests, all worked like I wanted them. One is the head for the compressor, the other one the fuel pump.


Time to have a beer ...
Nick

 

[mu38&Bg#]

At 108cc diesel injection is possible without making the injection system entirely too small. What is the original compression ratio? Will you increase it to account for the greater heat loss at scale size?

 

[MuellerNick]

I don't know what the original CR was. But with 108 cm^3, a CR of 20 should work.

Again, I intend to make a scale model. With the focus on scale.
If things don't work without disturbing the scale-ishm, they have to be hidden somewhere.
I have put a lot of effort into getting the right dimensions, and not just some guesswork with 3 pictures and thousand dimensions guessed. So I would be quite nuts to throw away that work (I'm nuts, but not stupid. Well most of the time).

Mill is just cutting its final paths for a new pattern, so hopefully two new castings this night ...


Nick

 

[canadianhorsepower]

AWsome work. are you going to put them in a tumbler after
are they will remain the way they are now

 

[MuellerNick]

are you going to put them in a tumbler after
are they will remain the way they are now

I will slightly glass-bead them. Helps to improve for the lacquer to stick on. I tried it with etch primer, but the coat is way too thick. I have 30% thinner in the lacquer, but it is a high quality one and a single layer is enough.

Casting #23. T-joint for coolant, fresh from today's first test pour.


Admittedly, I was a bit too violent when hammering out of the core sand mould (DTC) and it broke off (you can see the CA).


Core box and core. Takes two of them, part is symmetrical.

Well, lame! I can do better!

Casting #24. T-joint for intake. Also cast with DTC.


Out of the mold. Fail?


Not at all! Wall thickness in the pockets must be something less than 1 mm.
Edit: Way wrong! It is 2.5 mm!


Different view. The scratches in the bore are from removing the core.


And an other view, because I'm quite proud of it!


Core box and core for the mould. Also takes two of them.


Core box and core for the core. I'll add pins for aligning the two box halves.


Core in the core sand mould. Needs no fixing or aligning. Just drop it in.

I'll make a little modification to the core box of the mould to get more wall thickness at the gate. Just to be sure.


Well, can't complain!

I'll have a beer now …
Nick

 

[MuellerNick]

So you are lost in What-goes-where-city?
I do have a map for you:


Additions around the cylinder head. The second cylinder head will be left of part #24, just symmetrical.
#23 is roughly placed where it will be.


The A-frame with a few parts. The compressor is already known as a sub-assambly. But at that one, I had a wrong part.


And an other view, also with a Deckel FP1 mill (same scale) for size comparison.


Things will slow down a bit, as I have more patterns to make. I'm working at the gear-housing for the cam shaft (3 castings) plus the cam shaft mount that also serves as a mount for the cam shaft gear. Will be at about symmetrical to "/7" in the first picture of this post.

But don't worry, I have enough old castings to fill the gap!


Nick

 

[MuellerNick]

Got some news ...

Helped a friend to build a barn in the last few days. Scaffolding, filling concrete, etc. He told me, he is building a foundry. That was a lie! Now I tell everybody, he is building a Mosque. In the catholic Bavaria!
After that, I caught a flu. Hope to be back tomorrow or the day after.

But I got enough material to show. So two "new old" castings:

Casting #25. Generator housing, upper half.


The casting.


The pattern.


And the pattern (a split one) from the other side. This side goes up when casting.
One pattern part (the left one) in place, the other one sitting left of the index pins' bores.


The core. Where my indicator finger points to, there is a recess in the core. that shapes the oval mounting surface for the generator's collector ring arms (already shown).


The core box. The middle part has those pockets. The lower one is facing the wrong direction, it will close the mold.

This core is also used for the next casting ...

 

[MuellerNick]

Casting #26. Generator housing lower half.
That one is a bit more complicated.


The casting.
I had to bore two holes into the core (same as above casting) to insert wood screws into the core so I could lower it into the mould. These are the remains.


Detailed view. There is a pocket where the arrow points to. Also note, that my thump also slips into a pocket. That was necessary because the original was cast that way and to avoid shrinkage (that's why they did it).


Look into the pocked my thump slipped into.


The pattern. The face facing downwards looks the same like the upper half, also with two parts to go into the cope.


That's where one core goes to (black = core print).
This one makes the upper pocket.
The one for the other side is not the same, but mirrored.


Other core that makes the lower pocket (with my thump).
Same for the left side.
The two red lines meet in the mould.


Core box for the upper core (upper pocket). This one got a bit complicated. By exchanging two parts (in the upper left & right), it makes either the left or right core.


Partially disassembled ...

And completely disassembled.


Core box and core for the lower pocket (thump ...).

 

[MuellerNick]

And one picture missing ...

Core box for the lower pocket.
These two core boxes look quite odd. But the problem was, that the draft was either inward or outward. In this picture, part 3.1 was outward, and parts 3.2 and 3.3 were inward (or the other way round.

Here's a video, that might make it more clear how parts match ...
[ame="http://www.youtube.com/watch?v=L_Ppd_LJ5tM"]YouTube[/ame].

Edit:
That was my old core sand mixer. I have built a vibratory one in the meantime:
[ame="http://www.youtube.com/watch?v=7dFqHGV12is"]YouTube[/ame]


Nick

 

[MuellerNick]

That stupid not-a-foundry-barn is almost ready now, at least for this winter to come. And I'm way behind my schedule.

So a new old casting and some assembling (with hot glue).

Casting #27. Timing ring for the generator.


It takes two of them to make a ring (I think). Where I marked with the felt tip, that part gets milled off, just the lever is left. I did it that way, because there needs to be a groove around the circumference. In that groove, rolls go in that allow the ring only to rotate.
The two rings are bolted together in that small pocket near the end of the arc.


The pattern, a simple one.

And two assemblies:

The timing ring together with the arms for the brushes (to the collector).


And the generator housing with two of the arms that guide the timing ring.
The housing is about 230 mm wide, the inner diameter is 130 mm (IIRC).
The two blocks the housing stands on are ground level.


and the previous two sub-assemblies held together.


Nick

 

[MuellerNick]

... and a similar view of the original:

The lever for the timing ring to adjust is in the lower left, almost disappearing.
Someone has taken that one off, but I do have a drawing how it should look like.


Nick

 

[stevehuckss396]

outstanding!!!

 

[starnovice]

WOW Nick!

I have been doing castings for years and I just realized that I am still in kindergarten. You sure have given me some new directions to go.

What kind of glue do you use for your cores, especially when you glued them together for you "desk top casting"?

Thank you for making a little less ignorant than I was yesterday.

Pat

 

[MuellerNick]

Initially, I glued the cores with high viscosity super glue. That works, but not too well.
After having tried some different glues, I like the brand "Kolper 7" the most. That is a quick glue for cores. All core glues I have tried are water glass based.


Nick

 

[Dave Sohlstrom]

Nick

We need an update. I know I am going into withdrawal.

I tried to find information on Kolper 7 but came up with nothing.

Dave

 

[aonemarine]

Nick

We need an update. I know I am going into withdrawal.

I tried to find information on Kolper 7 but came up with nothing.

Dave

Me too
Its sunday and I have to go to work, an update would cheer me up a bit while there.

 

[MuellerNick]

Hmmm ... update!?
I do have made 3 new core sand molds. One worked perfect, one was a close miss and the third one a complete miss. Then, I have made three more patterns with each having a core. But I didn't cast these as I had to catch up with other things.

I think you'll have to wait until Tuesday for updates ... I'm sorry. But also happy that you are still reading.

Re the core sand glue:
Well, it looks like I have something top secret. I couldn't find it either.
It is best if you contact your local foundry equipment company and ask for a quick core sand glue. The one I use takes less than 1/2 hour until you can cast. There certainly is more than just one manufacturer.


Nick