Sprayed-on 3D Printed Pattern Filler

[GreenTwin]

Here is an info page for the Prusa slicer.

I am going to have to start taking notes, and documenting which settings I have used on the prints I have made so far, so I can get a comparison between print times, smoothness, etc.

https://help.prusa3d.com/article/layers-and-perimeters_1748

 

[Jasonb]

As I said earlier Pat, the OD of a flywheel rim is the worst place to hold one by so why make it flat for that purpose? You want access to the full width of the rim so it can be machined. Not least you need a machine that can swing another 3-4 inches so that the chuck jaws clear the bed, And using a chuck puts the work too far from the headstock to be able to make use of gap bed lathes.

Also your comment about clamping the side of the rim to a backerboard or faceplate is also not good as it means putting pressure on the spokes to hold it against the backer. a donut shape is better so the spokes bear on that and can be firmly clamped without risk of bending, distortion or even snapping the spokes.

 

[GreenTwin]

As I said earlier Pat, the OD of a flywheel rim is the worst place to hold one by so why make it flat for that purpose? You want access to the full width of the rim so it can be machined. Not least you need a machine that can swing another 3-4 inches so that the chuck jaws clear the bed, And using a chuck puts the work too far from the headstock to be able to make use of gap bed lathes.

Also your comment about clamping the side of the rim to a backerboard or faceplate is also not good as it means putting pressure on the spokes to hold it against the backer. a donut shape is better so the spokes bear on that and can be firmly clamped without risk of bending, distortion or even snapping the spokes.

Yes, I agree, the logistics of machining a 5-spoke flywheel will be a bit tricky.

I guess a 3D printed backer could be used in strategic places.

When I was machining the green twin flywheel, I did machine 1/2 the rim at a time, and ended up with a line in the center.
I eventually mounted it on a mandrel and made a light pass across the entire outside face, to get it completely flat.

I am not flywheel or engine expert by any measure, and so I will have to contemplate exactly how to do it, considering your comments, and find a lathe big enough to do it on.

It will be a good learning experience.

Edit:
I have not totally ruled out machining it in my mill, but that is a longshot in my opinion.
Machining it on my mill would keep all the machining in-house though.
I am thinking maybe use a carbide bit in a tool and die grinder, strapped onto the side of the mill somehow.
Or use a belt sander with the flywheel mounted on the rotary table on the mill.
The amount of material to be removed from a net flywheel is very small, if done correctly.

.

 

[ajoeiam]

As I said earlier Pat, the OD of a flywheel rim is the worst place to hold one by so why make it flat for that purpose? You want access to the full width of the rim so it can be machined. Not least you need a machine that can swing another 3-4 inches so that the chuck jaws clear the bed, And using a chuck puts the work too far from the headstock to be able to make use of gap bed lathes.

snip

Re: work to far from headstock - - -
Just use your heavy duty boring bar fastened on the rear side of the square toolpost (or similar positioning on other types).
That should give you a lot of reach AND give you a maximum diameter machinable.

 

[wce4]

I was playing around with various flywheel types/sizes/spoke configurations a few years ago (2019 to be exact), and 3D printed them in sections on my smaller Prusa.
I had forgotten that one of these was an early Ball Hopper Monitor prototype (the one on the right), and so this is how I split it.

I am not sure if these are optimum split line locations, but this is what I came up with in 2019.
This gives a 4-piece print (for one side) for a 5-spoke flywheel design.

If I want a true as-machined 14" diameter flywheel for the 1/2 scale Monitor, I will have to do something similar to this on the Prusa XL.

.
View attachment 152513

Thank you GreenTwin for all the info. One more thing would you be willing to share your stl file of this flywheel ? me likes a lot

 

[GreenTwin]

Thank you GreenTwin for all the info. One more thing would you be willing to share your stl file of this flywheel ? me likes a lot

I made the 3D file for that printed 5-spoke flywheel, and I am not sure if I can find it or not.
When I looked the other day, the one I found did not look exactly like that, but I will look again.

I made another one, larger, so between them, I will come up with something.

My 3D machine is at work, so I will have to wait until I return to make an stl file.

.

 

[DaleW]

I use a graphite and kerosene mix (used to be a product called plumbago) as a release agent on most of my patterns (unless casting steel). One of the great advantages is it fills in the layer lines on 3d printed patterns. So far it has worked in all sand types we use. I also have the luxury of using a chem set sand that we can control the setting time of, the slower the setting the better the finish on the surface (sometimes like glass). We treat our 3d printed patterns the same as any pattern as far as draft angle and contraction (I do a lot of reverse engineering of old components) we always try for a minimum of 3 degrees of draft on all straight vertical surfaces, always add for machining and contraction.

I always love reading about ways people have overcome obstacles on here, I have learnt so much and I thank you all for that, as I teach all my students there is always more than one way to do something, someways are just better than others and don't be afraid to experiment (just not on your bosses $30000 dollar project - maybe ask first)

 

[GreenTwin]

For casting work, I researched John Cambell's 10 rules for good castings, and also Bob Puhakka's (John Campbell's mentor) implementation of John Cambell's rules in a large scale industrial production environment, with huge industry-changing successes.

Opinions are mixed in the backyard community about John Campbell and Bob Puhakka, mainly from the armchair critics who don't do castings on an industrial scale, or perhaps who have never made a casting.
I focus on the quality of the castings made by folks, and don't get sidetracked by personality issues.
My motto is "Show me your castings", and lets go from there about discussions concerning methods.

When I started learning how to operate a foundry in 2011, I was advised to build a cupola/cupolet.
The lack of availability of coke for fuel eliminated this option.

I researched oil burners, and discovered that gray iron can be readily melted with this type burner.

I was told green sand / petrobond was really the only viable option for a backyard foundry, and was told that it would take me many years to figure out how to make quality greensand.
I ran across some local art-iron folks who used resin-bound sand, and fine commercial foundry sand (OK85), and so overnight I sidestepped any sand problems permanently.

And basically the entire backyard casting community uses traditional wood or aluminum flasks, which are very tall, and designed for use with greensand.
I discovered that resin-bound sand molds can be very thin, as thin as 1" or less, and can be shaped tightly around a pattern, to minimize sand usage (resin-bound sand cannot be easily reused).
And I discovered that resin-bound sand is sufficiently strong to stand by itself when poured, and so the universal backyard habit of spilling molten metal on your wood flask during a pour and burning the flasks can be eliminated, since you can use a wood flask with resin-bound sand as a removable snap flask.

So basically almost everything that I have been told from the backyard casting community has either been outright wrong, or not applicable to molds made from resin-bound sand.

For small engine parts, the resin-bound sand with a ceramic mold coat will approach investment casting quality, and yes, I can get a glass-like finish on gray iron castings right out of the mold, which no greensand mold that I have ever seen is capable of.
Having very dimensionally accurate engines castings really makes the machining process much easier, and gives a much more accurate engine.

I read the foundry industry publications, and I like to keep up with the latest trends, such as 3D printed very complex sand molds, net castings which are near or exactly at final dimensions right out of the mold, ceramic mold coats, ductile iron applications as replacements for weldments, etc.
If I see a promising method in industry, I will adapt that method if I can do it cost-effectively.

I have been using approximately 3% draft angle on engine patterns, and for non-machined surfaces, there is no problem with this being a part of the 3D model.
For machined surfaces, quite a lot of iron is wasted sometimes with a 3% draft that has to be machined off, but more importantly, the skewed surface can make it difficult to grip in a lathe chuck, and can present a danger of slipping out of the chuck.

The armchair folks will almost universally say you cannot pull a pattern from a sand mold unless you use draft angle.
I have played around with no draft angle castings, and have found that with resin-bound sand, you can indeed sometimes get away with little or no draft angle.
Castings with no draft angle on the surfaces to be machined gives a really nice and accurate part that is very near net dimensions, with minimal waste of material, and very minimal machining. Often just a light skim of the machined surface is all that is needed, since the part has almost no distortion and is true and square.

A substitue for resin-bound sand is sodium silicate bound sand, and with a quality sand, such as OK85 or equivalent, I think you could rival the resin-bound sand quality, without having to deal with the chemical exposure issues that occur with resin-bound sand.
And sodium silicate is readily available at pottery stores, and I think high quality sand is also available from the same stores.

I know a guy in the casting business, and he does not want to use bound sand because that would be wasteful, since it is not easily reused, and so he generally makes at least 4 molds, and 3 failures, and a final casting that is generally substandard, with a substandard surface finish.
There is no logic to this.

So if you really want to make high quality model engine castings, especially gray iron castings, it is best to ignore the "this is how it has always been done" and "this is how everyone does it" crowd, because there are far better methods and materials.

It is entirely possible to make very high grade gray iron castings in a backyard setting, and these castings can be defect free every time.
More than a few people have told me that this is impossible, but if you pay attention to what Bob Puhakka does, commercial-grade iron castings in a backyard setting is entirely possible.
I have gray iron castings to prove this; it is not just idle talk or wishful thinking.
No hard spots, no inclusions, no shrinkage defects, easily machined even when thin, and without annealing; clean smooth surface finish right out of the mold.
And I am moving towards an approximate "net" process, which means little or no draft angle on machined surfaces, and a casting that is amost the same size as the final machined part. I believe the net process can be achieved without much trouble in a backyard foundry setting.

.

 

[GreenTwin]

There are lots of people in the backyard casting community who spend a lot of time designing pour basins, and so it is generally accepted that a pour basin is required for good castings.

I am of the opinion that if you want lots of defects in your castings, use the traditional pour basin, and that by itself will almost guarantee you will have defective castings.

Rule #1 (in my mind anyway, and according to Bob Puhakka) is to control the velocity of the metal at all times, and also don't ever let the metal splash and turn back onto itself.
You will entrain air, slag, sand, etc. if you have excessive velocity, and if you splash the metal around, such as occurs when using a pour basin.
Its the worst thing ever invented in the casting world.

The challenge is to break the mindset of the backyard casting community about what is standard accepted foundry molding practice, and what actually is required to produce defect-free castings.

Making defect-free castings is the easy part.
Getting people to change their mindset about how to make castings is what is nearly impossible.
It is the proverbial "You can lead a horse to water......" thing.

There is nothing magical about what I am doing to make commercial-grade iron castings; I am just following the rules and methods of what other highly successful folks are doing to make high grade defect-free castings.

.

 

[Jasonb]

When you started this drive to make the old casting sets available to all again you wanted to put the patterns where others could make use of them. To me it would make sense to have the patterns where a foundry would be willing to cast them and the costs not be prohibitive. If the patterns can only be cast in resin bound sand and time has to be taken lining edges with Mylar then that will drive up costs and limit who will want to take on casting. Also your use of hollow patterns to produce the cores as you don't like making core boxes will again drive up costs and limit foundtries that will take on the job as it is more labour intensive than just using a corebox.

Might be OK for someone just doing their own thing in the back yard but very few have the ability particularly in iron so to me it would be better to have more universal patterns.


Maybe Dale could comment on the likely cost implications of using a traditional pattern with draft etc that can be cast in green or airset against one that has no draft, needs lining with mylar, uses the pattern to act as a core box, needs resinset, etc. I would say the foundry that is casting the parts I'm making patterns for would be double the cost in many cases reducing slightly as the size of the casting increases

 

[Steamchick]

Hi Green Twin,
Just picked up on this thread... so probably repeating what others suggest...
Excuse my ignorance.
Your note "since the rim should be pretty flat, and thus would give a flat surface to pull up to." ---- Well "pretty" it may be, but we are really talking of a "proper job" - as you always do. So "pretty" is not what I call "precise".
You need to:

1. Have a stiff enough face-plate on the lathe so the face-plate does not distort when tightening the clamps to hold the flywheel. And then if you clamp a cast surface to the machined face-plate you are very likely to distort the casting to be a shape you don't want when you un-clamp it and it springs-back.... Always clamp castings to shimmed spacers, so there are NO distorting stresses generated in the casting while it is being machined..
2. You MUST allow for the tool to cut right past the casting if machining the outer face - Or it ends-up looking "Horrid". Not like your other engines. - So that means spacing off the surface of the face-plate. So clamp each spoke to a shimmed spacer to get the centre as true as you can, then follow the next step, which can avoid the need for the face-plate for machining the outer rim, unless the flywheel is a "Light" (scale?) design and not stiff enough to manage being machined when unsupported.
3. MY preferred method for flywheels - (but my largest was 6in dia (brass) from a bought casting). MACHINE THE BORE AND BOSS FIRST. Then mount the flywheel on a mandrel (Clock to be sure it is true). Only then machine the outer face and sides of the flywheel - if they need to be machined for a belt drive or some such. A pretty good casting may just need flash and sprues tidying-up with a file, then balancing to counterbalance the con-rod and piston. - Which will save buying a new lathe with >16-inch swing!
4. If you really need the outer rime machining, and NOT looking wobbly, you must have room on the stroke of the cross-slide and tool mounting to be able to get the tool outside the OD of the flywheel - and be STIFF enough not to simply follow every contour of the casting. Also, you should cut both faces and the outer diameter in a single setting if you really want the flywheel to look good. I have seen flywheels machined on faces so when running the outer rim varies in width... visibly.. and looks horrid.
5. And then the lathe must be capable of handling torque at a low-enough rotational speed for the tool to work properly cutting the hard skin of the casting.... Variable speed lathes don't do this well in my experience. But old lathes that had back-gearing will do this well. (picture of a Myford ML3 that I used for a 6in flywheel - (Back gear adjacent to the headstock beneath the cover that caught the surplus grease from the gears!).
   
6. Undercutting the inside of the rim often simply looks eccentric, if not done on the same set-up as side faces and outer... And is it necessary? The "sharp" corners some people use, and cut into spokes, just do not look "authentic" to me, as the Victorians would avoid machining anything if they could - aware of costs not "pretty", and it is also a way of showing-off the quality of the castings.... - and selling the skills of your foundry.

Just my humble opinion, as you do produce really excellent work, that I admire.
K2

 

[GreenTwin]

I was reading a recent discussion about Bob Puhakka and John Campbell's methods, and the question was asked "If those methods are so good, then why doesn't everyone in the foundry industry use them ?".

This reminds me of a similar question that was asked by anyone who started using one of Charles Porter's new "high speed; ie: 300 rpm) steam engines, which was "Why doesn't everyone use one of these engine types ?".

It did not take very long for the entire world to change to the high speed type stationary steam engine.
The folks who did not adapt to this modern technology generally went out of business.

.

 

[GreenTwin]

Hi Green Twin,
Just picked up on this thread... so probably repeating what others suggest...
Excuse my ignorance.
Your note "since the rim should be pretty flat, and thus would give a flat surface to pull up to." ---- Well "pretty" it may be, but we are really talking of a "proper job" - as you always do. So "pretty" is not what I call "precise".
You need to:

1. Have a stiff enough face-plate on the lathe so the face-plate does not distort when tightening the clamps to hold the flywheel. And then if you clamp a cast surface to the machined face-plate you are very likely to distort the casting to be a shape you don't want when you un-clamp it and it springs-back.... Always clamp castings to shimmed spacers, so there are NO distorting stresses generated in the casting while it is being machined..
2. You MUST allow for the tool to cut right past the casting if machining the outer face - Or it ends-up looking "Horrid". Not like your other engines. - So that means spacing off the surface of the face-plate. So clamp each spoke to a shimmed spacer to get the centre as true as you can, then follow the next step, which can avoid the need for the face-plate for machining the outer rim, unless the flywheel is a "Light" (scale?) design and not stiff enough to manage being machined when unsupported.
3. MY preferred method for flywheels - (but my largest was 6in dia (brass) from a bought casting). MACHINE THE BORE AND BOSS FIRST. Then mount the flywheel on a mandrel (Clock to be sure it is true). Only then machine the outer face and sides of the flywheel - if they need to be machined for a belt drive or some such. A pretty good casting may just need flash and sprues tidying-up with a file, then balancing to counterbalance the con-rod and piston. - Which will save buying a new lathe with >16-inch swing!
4. If you really need the outer rime machining, and NOT looking wobbly, you must have room on the stroke of the cross-slide and tool mounting to be able to get the tool outside the OD of the flywheel - and be STIFF enough not to simply follow every contour of the casting. Also, you should cut both faces and the outer diameter in a single setting if you really want the flywheel to look good. I have seen flywheels machined on faces so when running the outer rim varies in width... visibly.. and looks horrid.
5. And then the lathe must be capable of handling torque at a low-enough rotational speed for the tool to work properly cutting the hard skin of the casting.... Variable speed lathes don't do this well in my experience. But old lathes that had back-gearing will do this well. (picture of a Myford ML3 that I used for a 6in flywheel - (Back gear adjacent to the headstock beneath the cover that caught the surplus grease from the gears!). View attachment 152537
6. Undercutting the inside of the rim often simply looks eccentric, if not done on the same set-up as side faces and outer... And is it necessary? The "sharp" corners some people use, and cut into spokes, just do not look "authentic" to me, as the Victorians would avoid machining anything if they could - aware of costs not "pretty", and it is also a way of showing-off the quality of the castings.... - and selling the skills of your foundry.

Just my humble opinion, as you do produce really excellent work, that I admire.
K2

All excellent points in my opinion.

.

 

[GreenTwin]

When you started this drive to make the old casting sets available to all again you wanted to put the patterns where others could make use of them. To me it would make sense to have the patterns where a foundry would be willing to cast them and the costs not be prohibitive. If the patterns can only be cast in resin bound sand and time has to be taken lining edges with Mylar then that will drive up costs and limit who will want to take on casting. Also your use of hollow patterns to produce the cores as you don't like making core boxes will again drive up costs and limit foundtries that will take on the job as it is more labour intensive than just using a corebox.

Might be OK for someone just doing their own thing in the back yard but very few have the ability particularly in iron so to me it would be better to have more universal patterns.


Maybe Dale could comment on the likely cost implications of using a traditional pattern with draft etc that can be cast in green or airset against one that has no draft, needs lining with mylar, uses the pattern to act as a core box, needs resinset, etc. I would say the foundry that is casting the parts I'm making patterns for would be double the cost in many cases reducing slightly as the size of the casting increases

This hobby had changed a LOT in the short time I have been involved in it in a serious way (I started studying model engine design in 2006 after my dad died).

Many folks were making hand drawings, and 3D modeling was considered by many to be an exotic and generally unnecessary method for typical model making, and beyond the capability of most hobby folks.

Fast forward to 2023 and you seen many folks using 3D modeling on a hobby setting with fantastic results.
3D modeling is no longer seen as something exotic, but instead something very useful, and cost effective too given the accuracy that can be achieved with it, and the affordability of modern 3D software.
The quality of the handmade drawings from many of the old casting kits often left much to be desired, and these drawings generally needed multiple corrections.

3D printing was really not a thing in the hobby just a few short years ago, and now it is a very big deal as 3D printer prices come down, and more folks become proficient at 3D modeling.

CNC machining has also become a big deal, and has gained pretty widespread use as best I can tell.
Look no further than JasonB's CNC'ed pattern work to find out just what a magnificent pattern looks like.

So it seems like the argument being forwarded is that even though we can make great technological advances in the hobby with 3D modeling, printing, CNC, we should not expect any advances on the hobby foundry side, or even advances on the commercial foundry side.

I guess it begs the question "Can a hobby person find and use a commercial foundry that will produce affordable castings ?".
And it seems that the assumption is that an affordable commercial foundry that can be used by the hobby must use greensand and greensand techniques only.

This is a question of no doubt much debate, and I am not sure there is a simple answer to it.

For someone who wants to create a full-blown casting company (which does not include me), then standard matchplate patterns, coreboxes, etc. would pretty much be a must.

I have been told recently that one follower of the Ball Hopper Monitor thread knows of a foundry that will do limited runs with loose patterns, and so I know those foundries exist.

For a commercial production run of any quantity, I think the methods that Maury used (Lone Star casting kit guy) would be necessary; ie: use Freeman foundry products to cast hard plastic resin copies of patterns, and similarly cast plastic negatives to create coreboxes.
Maury said you should never send original patterns to a foundry, and I have read horror stories of wood motorcycle engine patterns that took thousands of hours to produce that got left near an open window, and got wet and ruined.

I really don't see high number production runs of any casting kit of any size being made these days; I think it will be at best a limited production affair at a commercial foundry.

The key to making limited quantities of castings in a backyard setting is to get a perfect casting with each mold, to minimize waste, time and materials.

So I really don't want to compromise my casting quality so that others can make what I consider substandard castings.
Not trying to be snobbish, just saying.
I am trying to promote the model engine hobby in general, and especially to promote the backyard-made engine castings, and the art of doing that.

One of the problems is that STL and similar 3D files are not easily modified when shared across various 3D programs.
If one could easily modify an STL or similar file with any 3D program, then it would be easy to add any draft you wanted to have.

I don't have all the answers at this point, but I will be making some castings for the Ball Hopper Monitor, and I hope that they are near-net.

.

 

[GreenTwin]

This is a Freeman video, and while I don't understand every use for their large selection of castable platic materials, I do get the general idea about how you can make both hard-rigid plastic castings, and flexible rubber-like castings.

Freeman sells some very versatile products, and you can do a lot with them, assuming you watch their videos and learn how to do it.

For someone really serious about making the patterns to send to a foundry for mass production, this is one method I have seen used.

.

 

[GreenTwin]

Some hard and flexible Freeman materials.

.

 

[GreenTwin]

Here is an interesting one.

Check out the resin-cast distributor cap at 4:00.
Now how cool is that ?

.

 

[GreenTwin]

When I was trying to make the green twin castings, with the intent of making a couple of sets of castings for a two-person group build, I naively thought that one could just place most or all of the patterns in one large mold.

I had seen others do this, perhaps in not as large of a flask.

 

[GreenTwin]

I was trying to use a brute-force method with very tall sprue, to basically force the iron into the very long runner and mold cavities.

This method did not work well at all, partly because I had not yet learned how to correctly tune an oil burner, and mainly because there was too much cooling effect down the very long path that the molten metal had to take.

At the time I made this mold, I did not have a working knowlege of how a correctly functioning sprue/runner/riser system worked, and so I just blindly stumbling along.

I got a few usable parts out of this method, but in general, this turned out to be a very bad/wasteful method in my opinion, with lots of partially filled mold cavities.

.

 

[GreenTwin]

This was a later molding technique that I used, with a minimized flask that fit the pattern more tightly, and could be poured in snap-flask fashion.

This method worked well with aluminum 356, and I think would have worked well with iron had I understood at the time how to optimally tune my oil burner, and how exactly to melt and skim iron in a reducing atmosphere.

This method used a tall sprue, with a ceramic filter at the base.
I have gotten away from the ceramic filter, since it is not necessary if you use a spin trap at the end of each runner.
.