I am getting ready to try my hand at a sand casting: the 1/5 scale Chenery rotary Crankcase in aluminum. I am not convinced that I would need a ‘split pattern’ for this casting. Are there any suggestions on where to sprue / gate pattern this mold? Thanks, Norman
Aluminum sand casting for Chenery Gnome Rotary
- Thread starter albd5a
- Start date
Help Support Home Model Engine Machinist Forum:
Opinions on how to approach making a casting for the Chenery Gnome Rotary will vary.
I can offer one way to do it.
A photo of what I think the casting would look like is at post #22 here:
https://www.homemodelenginemachinist.com/threads/les-chenery-rotary-engine-1-5.31718/page-2
Assuming the casting is laying flat on its back, with the crankshaft facing straight upwards:
It probably would be easier to split the pattern hozirontally at the centerline of the cylinders, but you can always use a follower board with a one-piece pattern.
I think I would use a 3-piece flask, and let the runner be in the lowest flask.
The runner would be horseshoe-shaped, with a vertical sprue at the center of the horseshoe.
Two long knife gates from the runner, one each side, into the bottom/side of the crankcase.
You could use spin traps at the ends of the runners if desired.
The core in the center would be round, with radial coreprints going out for the cylinder holes.
My 3D model machine is not set up at the moment, so I can't sketch it up.
Extend the bottom of the casting down to allow machining allowance, and to allow the gates to enter the mold cavity.
Then machine the gates off.
You could add a boss on the top of the casting to allow you to grip it, and then machine that boss off.
I would consider using a sodium-silicate bound mold for the cores and mold, since this part seems to be relatively small.
.
I can offer one way to do it.
A photo of what I think the casting would look like is at post #22 here:
https://www.homemodelenginemachinist.com/threads/les-chenery-rotary-engine-1-5.31718/page-2
Assuming the casting is laying flat on its back, with the crankshaft facing straight upwards:
It probably would be easier to split the pattern hozirontally at the centerline of the cylinders, but you can always use a follower board with a one-piece pattern.
I think I would use a 3-piece flask, and let the runner be in the lowest flask.
The runner would be horseshoe-shaped, with a vertical sprue at the center of the horseshoe.
Two long knife gates from the runner, one each side, into the bottom/side of the crankcase.
You could use spin traps at the ends of the runners if desired.
The core in the center would be round, with radial coreprints going out for the cylinder holes.
My 3D model machine is not set up at the moment, so I can't sketch it up.
Extend the bottom of the casting down to allow machining allowance, and to allow the gates to enter the mold cavity.
Then machine the gates off.
You could add a boss on the top of the casting to allow you to grip it, and then machine that boss off.
I would consider using a sodium-silicate bound mold for the cores and mold, since this part seems to be relatively small.
.
I'm not sure how you can cast it without a split mould unless you are going to investment cast. Pat's link on page 2 looks to show a crankcase cut from solid so it does not have the round bosses that the cylinders screw into or the webs between them.
Just don't feed it via a "cast "surface and at such a small size I would not bother with cores, you don't see it on others. Just a single hole through the middle which does not need a core
There is one up for sale at the moment on ebay and also on MECH forum which may save you the bother
Just don't feed it via a "cast "surface and at such a small size I would not bother with cores, you don't see it on others. Just a single hole through the middle which does not need a core
There is one up for sale at the moment on ebay and also on MECH forum which may save you the bother
Attachments
It might be easier to make as a single piece pattern and an oddside board. I also have the possibly mistaken idea that an oddside helps with
registering the mould boxes accurately.
registering the mould boxes accurately.
I have made bound sand pattern halves, and I use snap flasks.
So once the flask is removed, there needs to be something to index the cope with the drag exactly.
There are several ways to do the indexing, but the way I do it is before I remove the flasks, but after the two mold haves have hardened, I drill two 1/4" holes in the mold with a 12" long drill bit, and use 1/4" wood dowls inserted in the holes to get the mold halves to fit back together perfectly.
If you are using a follower board, and green sand, the follower board does not have to register to anything, but when you remove the follower board, and add the 2nd flask half, that flask half must index to the first flask half with dowls/pins.
The follower board just has to be larger than or the same size as the flask.
.
So once the flask is removed, there needs to be something to index the cope with the drag exactly.
There are several ways to do the indexing, but the way I do it is before I remove the flasks, but after the two mold haves have hardened, I drill two 1/4" holes in the mold with a 12" long drill bit, and use 1/4" wood dowls inserted in the holes to get the mold halves to fit back together perfectly.
If you are using a follower board, and green sand, the follower board does not have to register to anything, but when you remove the follower board, and add the 2nd flask half, that flask half must index to the first flask half with dowls/pins.
The follower board just has to be larger than or the same size as the flask.
.
If using woodern patterns then a follower board would be considerably harder to make than the actual pattern. Easier if you are 3D printing as it is just a simple boolean subtraction.
Though if in the hole foundry I suppose you could simply scoop away the sand down to the mid line.
Though if in the hole foundry I suppose you could simply scoop away the sand down to the mid line.
![MeshMagic 3D Free 3D Modeling Software [Download]](https://m.media-amazon.com/images/I/B1U+p8ewjGS._SL500_.png)
If you put a pattern in a flask, and fill the flask, ram the sand, etc., then you should flip the flask, and scoop down to the partline from the backside, since the chances of getting good compaction on the sand below the pattern are not that good.
The Navy Foundry Manual decribes a "False Cope Method", where an irregular part is rough molded to make a temporary cope, then a permanent drag mold is made, and finally a permanent cope mold made.
Good for irregular shapes, broken parts that are used for patterns, etc.
One use of this method is to make molds for boat propellers.
.
The Navy Foundry Manual decribes a "False Cope Method", where an irregular part is rough molded to make a temporary cope, then a permanent drag mold is made, and finally a permanent cope mold made.
Good for irregular shapes, broken parts that are used for patterns, etc.
One use of this method is to make molds for boat propellers.
.
Last edited:
As I said in the other thread unless you have a 3D printer there will be a lot of work making a follower for this part. The actual pattern is very simple in wood but to get a good fitting follower will be much more work.
The ring and it's web can be turned from two flat pieces of wood glued together with a sheet of copy paper between. Then bore 9 radial holes and glue in your round pieces which can again be turned as a long length with paper on the split line and then cut to length. Bondo the fillets and once sanded just split it apart along the paper line.
If you want a load more work then make a follower by carving this out of wood, I've not shown the radii needed to clear the fillets on the pattern.
To me for a one off it is a wast of time making a follower unless you can print it, even then is there any point?
The ring and it's web can be turned from two flat pieces of wood glued together with a sheet of copy paper between. Then bore 9 radial holes and glue in your round pieces which can again be turned as a long length with paper on the split line and then cut to length. Bondo the fillets and once sanded just split it apart along the paper line.
If you want a load more work then make a follower by carving this out of wood, I've not shown the radii needed to clear the fillets on the pattern.
To me for a one off it is a wast of time making a follower unless you can print it, even then is there any point?
You don't need all that. You only need what is 'seen' at the parting line. Just an annular recess and some shaped dowel like pieces let in to the periphery for the spaces between the cylinders. I would draw it, but I don't have time right now.
Not sure what a follower is. It’s not in any of the books that I have on casting it seems that it would act as a temporary, cope, thereby keeping that 1/2 of the pattern exposed free of sand. Then you roll over the drag, remove the follower and create a real cope? I watched a video of the Windy Hill foundry where he casts a similar structure, namely, a large wheel, and that is not a split pattern and he does a lot of tweaking by moving the sand manually here and there…
Thats basically it the follower stops the sand at the split line like you were moulding the first half of a split pattern
What are you making your pattern from? wood or 3D printing
What are you making your pattern from? wood or 3D printing
A follower is just something to support the backside of a 1-piece pattern, so you can ram up to the follower, which is at the parting line.
A follower can be a board with a hole in it, like Jason shows above, or it can be 1/2 the pattern set into plaster, bondo, clay, or anything that will support the pattern and allow you to ram up to the parting line.
A modern follower could be 3D printed, as Jason mentions.
Once you ram up to the parting line, remove the follower, flip the flask, and ram the 2nd side.
The info below is from the Navy Foundry Manual.
See top-right paragraph, page 67 for follower.
.
A follower can be a board with a hole in it, like Jason shows above, or it can be 1/2 the pattern set into plaster, bondo, clay, or anything that will support the pattern and allow you to ram up to the parting line.
A modern follower could be 3D printed, as Jason mentions.
Once you ram up to the parting line, remove the follower, flip the flask, and ram the 2nd side.
The info below is from the Navy Foundry Manual.
See top-right paragraph, page 67 for follower.
.
Last edited:
Thank you for your help. Thanks to everyone. My ‘pattern’ is wood with a lot of primer and sanding. I’ve just sprayed with gloss Rustoleum. Will wax with carnauba when dry…. If I split the pattern, not sure if I will get it perfect and ideally would have to compensate for the band saw Kerf. … I may or may not.
OK, thanks. I am quite sure that I now understand what a follower is. I’ve done that with Clay when making split silicone molds for resin, casting. I do have a ton of clay. I probably could’ve had it done by now rather than thinking about it. Starting to sound like splitting the pattern would be easier than doing the follower.
OK as you have made the pattern solid then you are best either with some form of follower or just scooping out the sand down to the mid line which can be marked onto the patern with a sharpie before you get in in the flask.
I would tend to make a wood pattern like that with a glue joint through the middle, piece of copy paper sandwiched in the joint.
You can then machine it as if it were solid
The once it is finished a quick blow to a chisel will split the glued joint apart. You can then just lightly sand to remove th eremains of the paper and have two halves that match correctly.
I would tend to make a wood pattern like that with a glue joint through the middle, piece of copy paper sandwiched in the joint.
You can then machine it as if it were solid
The once it is finished a quick blow to a chisel will split the glued joint apart. You can then just lightly sand to remove th eremains of the paper and have two halves that match correctly.
There is another important use of a board. Sometimes it is not possible to have a parting line that lies in a single plane, at least without using complicated cores. This can give the moulder a lot of skilled sand carving to do. Here is a one-piece bracket pattern that illustrates the point.
The solution is a shaped follower board called an "oddside." Actually, oddside is the term I use for any follower board, flat or bumpy, but I may be wrong there.
One side is rammed up with the pattern sitting in the oddside. Then the box is turned over and the oddside removed, leaving the pattern in place. Moulding the second box fills the cavity left by the oddside, forming a matching, contoured parting line. Note the generous draw and radii on the outside of the oddside, making the projecting lump of sand on the second box less vulnerable.
Would the first box to be moulded be the cope or drag? I don't know.
The solution is a shaped follower board called an "oddside." Actually, oddside is the term I use for any follower board, flat or bumpy, but I may be wrong there.
One side is rammed up with the pattern sitting in the oddside. Then the box is turned over and the oddside removed, leaving the pattern in place. Moulding the second box fills the cavity left by the oddside, forming a matching, contoured parting line. Note the generous draw and radii on the outside of the oddside, making the projecting lump of sand on the second box less vulnerable.
Would the first box to be moulded be the cope or drag? I don't know.
Last edited:
I decided to split my wood pattern on the bandsaw; I am glad I did. Obtain perfect alignment of the two halves with steel pins references. The difficult part was getting the inside of the crank case hollow rather than one chunk of metal to be bored out later. ( did not use a sand-water glass Core. However, I was able to get enough draw on each half to get a casting. I never melted aluminum before so at least I figured out how to do that now. Sure enough , unfortunately I have a slight miscast -
That is a very nice casting !
Your riser located on the left has a very small gate into it, and so that gate solidified before the main casting fully solidified, and the remaining molten casting acted as the riser, and caused the shrinkage.
Perhaps use two risers at a 45 degree angle, and make the gate into the bottom of them much larger.
The gate could be perhaps six times larger than what you used.
You will machine that area anyway, so the size of the gate does not matter.
That is a really good casting for a first-time aluminum melt.
I tended to overheat my aluminum when I first started to cast metal, and that caused a very rough surface finish.
Now I use a pyrometer, and pour at approximately 1,350 F, which gives a good surface finish like you have.
Your mold alignment is very good.
This is a most excellent casting effort for sure.
.
Your riser located on the left has a very small gate into it, and so that gate solidified before the main casting fully solidified, and the remaining molten casting acted as the riser, and caused the shrinkage.
Perhaps use two risers at a 45 degree angle, and make the gate into the bottom of them much larger.
The gate could be perhaps six times larger than what you used.
You will machine that area anyway, so the size of the gate does not matter.
That is a really good casting for a first-time aluminum melt.
I tended to overheat my aluminum when I first started to cast metal, and that caused a very rough surface finish.
Now I use a pyrometer, and pour at approximately 1,350 F, which gives a good surface finish like you have.
Your mold alignment is very good.
This is a most excellent casting effort for sure.
.
