Ball Hopper Monitor - Casting Project

[GreenTwin]

I toyed around with the water hopper window, but can't really make it any better.

The fillet option for compounded curves does not work well at all in my version of Solidworks.

The wall thickness of the window cutout has to remain uniform.

Looking at the photos, I don't think I have my window profile quite right.

And it seems like a retract was used at the window, which perhaps pulled down.

Edit:

Looking at the photos again, I originally thought the hopper pulled sidways from either side, but on closer inspection of a photo of the window, it appears the hopper pulls up and down, which my window would actually do.

Edit2:

I think it will be wise to 3D print the cylinder, vavle chamber, and carb, and make sure everything will fit, and make sure the bolts and nuts will fit on.
Having a bolt or not not go on would be a show stopper if the tank ended up having to be recast.

Edit3:
The core will have to be split vertically to get a pull off of the window interior.
The exterior mold will have to be split horizontally at the centerline of the hopper.

.

 

[Jasonb]

I'd still add draft to the end of the filler boss, like I have highlighted here in green & Yellow, split is the horizontal mid plane. The curved surfaces are self drafting.

 

[Jasonb]

Split the hopper horizontally through the bead and it will all pull out OK as will the core, The inner face of my window has draft as well as the outside hence the "uniform thickness" so will pull easily with no retracts

Why not just assemble the parts in CAD and check them for fit, no need to print to try them out.

If you can't draw the fillets then just leave them square edged and sand the external corners and bondo the internal but kind of defeats the object of using the Tech to model and print the final patterns

 

[GreenTwin]

The ends of bosses only need a draft like that if you are pulling perpendicular to them, not if you are pulling axially, which is the case with the outlet on the bottom of the fuel tank.

I guess I could do an assembly in Solidworks, but sometimes that can be deceiving a bit.

I will have to fill some fillets inside and out on the window.
Does not defeat the purpose of 3D modeling/3D printing, but is a little extra effort.
The 3D modeling and 3D printing gives you a very consistent wall thickness through some very tricky parts and shapes, with is very difficult to duplicate with manual patterns.

Solidworks is not perfect, and my knowledge of it is limited, so if I can get close, then I can deal with a little filler on the 3D printed patterns, which is pretty normal for me.

Edit:
One of the biggest problems I have seen with folks making manual patterns is that the cores are not accurate, nor are the patterns, and so when complex castings with passages, such as I have seen on compound steam engine cylinders, are sectioned, the wall thickness is often not uniform, and there are often places where the wall thickness is unacceptably thin.

The variation in wall thickness causes uneven solidification.
One of the rules of good castings is to keep the thickness of the walls at a constant thickness, for even solidification.
If you look at a section of any old engine, generally this rule is adhered too.

3D modeling and 3D printing gives you a consistent wall thickness, and gives you accurate cores that are accurately positioned.


.

 

[lee webster]

When I am unsure if a pattern will pull cleanly from the sand, I make a sand mould in CAD, subtract the pattern from it, and then see if I can move the pattern out of the mould. All in CAD. It's helped me quite a bit.

 

[Jasonb]

Edit:
One of the biggest problems I have seen with folks making manual patterns is that the cores are not accurate, nor are the patterns, and so when complex castings with passages, such as I have seen on compound steam engine cylinders, are sectioned, the wall thickness is often not uniform, and there are often places where the wall thickness is unacceptably thin.

I think this is something that can't really be helped when scaling a model, take your hopper on the full size it may only have been say 3/8" thick but to scale it at 0.42:1 it is only going to be 5/32" thick on the model so beginning to get difficult to pour without the iron cooling before the mould has been filled completely.

If you then add in passages for steam, water or combustion gasses you don't want to restrict these so the model maker needs to keep them as large as possible which may well be over scale which further reduces wall thickness.

 

[GreenTwin]

I don't try to scale the wall thickness on thin castings; that generally does not work.

I add material on the inside, where it cannot be seen.

I think about 0.25" for wall thickness is as thin as I try to go, at least on this scale engine.

You can get away with a lot if you use phosphorus in the iron, such as old radiators, but I hope to avoid that, at least on the cylinder.

.

 

[GreenTwin]

If you get two areas adjacent to each other that are not the same thickness, you have to be careful or you will get hot tears.
Sometimes you cannot avoid unequal thicknesses, so you have to use various methods to deal with that.

The solidification needs to be progressive across the casting, and not not leave a hot spot between two or more areas that have solidified.

This is addressed in John Campbell's 10 rules for good castings.

The runner should solidify last, and the solidification process should proceed towards the runner.

http://pmt.usp.br/ACADEMIC/martoran/NotasFundicao/JOM-CampbellRules.pdf
.

 

[GreenTwin]

Water hopper lid is done.

This was one of the easier pieces to model.

I probably need to add a slight amount of draft angle on the lower lip, but I may try to pull it from the mold without draft angle, to basically get a net sized casting that only needs a little cleanup.

Resin-bound sand is somewhat tolerant of pulling with no draft angle, if the pull is not too deep.

I really need to add some draft angle on the inside as a minumum, since that could cause a lot of problems.

I normally don't include draft angle in the 2D drawings, since that becomes too confusing.

.

 

[Jasonb]

On a single thickness item likethe hopper thickening up is OK but for say passages between two compound cylinder that you mentioned the bore and ctr distance stays the same so you can't thicken things up without reducing the passages.

What direction are you going to pull that lid out of the mould?

 

[GreenTwin]

What direction are you going to pull that lid out of the mould?

The lid mold will have to be a three piece, with two top pieces pulling sidways, and the bottom piece pulling down.

.

 

[GreenTwin]

On a single thickness item likethe hopper thickening up is OK but for say passages between two compound cylinder that you mentioned the bore and ctr distance stays the same so you can't thicken things up without reducing the passages.

If I shorten the water passages by about 1/3, I can thicken the outside water jacket (inwards) to about what the rest of the wall thicknesses will be.

That is the game plan.

.

 

[GreenTwin]

The piston model.

.

 

[GreenTwin]

I would have sworn on a stack of Bibles that I had made a 3D model for the Ball Hopper Monitor with ellipsoidal spokes, but I cannot find that in my 3D model collection.

Best I can find is 3D model with flat spokes that have had fillet added.

Flat spokes will not do for this build, so I will have to revise or restart my flywheel 3D model for the flywheel.

The original old engine flywheel spokes were almost universally ellipsoidal in section, with a gracefull taper along their length.

It is quite difficult for me to get flywheel spokes to look visually like the old flywheels, but I have gotten much better at that.

.
These are my original BHM flywheel models, and a photo (not mine) that I used as a go-by.

.

Edit:
I suspect I changed the BHM spokes to ellipsoidal, and then lost that model somewhere, since the fillets on the 3D print on the right are slighly smaller than what is in my original flywheel 3D model.
I think I like a slightly larger fillet at the spoke/rim junction than is in the 3D print.

.

 

[GreenTwin]

The outer muffler half, in 3D.

.

 

[GreenTwin]

Cylinder head and crankshaft.

.

 

[GreenTwin]

Muffler, inner half.

I think the text should be rotated 90 degrees, but I could not get SW to do that.

.

 

[GreenTwin]

It is probably more like this, although I don't have a photo to indicate the exact position.

.

 

[GreenTwin]

Slowly working towards the all important frame/cylinder.

There are a awful lot of round/spherical shapes in this engine.

.

 

[GreenTwin]

It looks like several folks have used some creative methods to bore the cylinder for this engine.

One build had a boring bar mounted in the lathe chuck, with the engine frame mounted on an angle bracket on the lathe carriage.

Another build used a boring bar in the lathe chuck, with a bearing in the bottom of the frame that the boring bar ran through, and the frame bolted to the lathe carriage, but the end of the boring bar only supported by the bearing in the frame.

And Jason supported his lathe-chuck-mounted cylinder with a steady rest, and bored with a lathe boring bar.

The original full-sized Ball Hopper Monitor had a hole in the bottom of the crankcase, and no doubt this was used to accommodate a boring bar, since the hole is the same size as the bore of the engine.

My cylinder/frame will be 14.42" tall, so I may build some sort of bracket that bolts to the head end of the cylinder, and bolts to the lathe carriage, and then let the feet of the casting slide on the lathe bed rails, or slide on blocks of wood to protect the rails.

.