Ball Hopper Monitor - Casting Project

[GreenTwin]

One of my pet peeves (one of many unfortunatley, but the only one I will complain about today) is trying to get wonky casting back to some datum surfaces.

I machined some Myers steam engine castings, and I recall them being pretty straight and accurate, and requiring very little machining to get good surfaces.

And the Lone Star Ball Hopper monitor castings were 1st class and very accurate, and apparently were cast in bound sand of some type.

The castings I made for the green twin that were cast in bound sand had a high degree of roundness/flatness/straightness (for lack of the correct terms).

In the industry magazines, they call accurate castings "near net" as I recall.

So for the green twin, most parts required just a very light cut, and the flywheel was very round both on the outside of the rim, the face of the rim, and the inside surfaces of the rim.
It was very easy to get the raw casting centered on the lathe, and just a light skim here and there and the part was done.

I have gotten spoiled on parts that are near net, and so I am trying to do that with my future castings.

It would save a lot of time to have the bottom of the frame/cylinder feet flat, and so I am scheming about casting the top and bottom of the frame/cylinder with no draft angle at all.

They make these heavy sheets of mylar, to be used as covers for drafting tables (not the old-style covers that were very thick laminations, but rather a semi-thick mylar film, and I have some of that material.
I would guess it to be about the thickness of heavy card stock.

So the plan is to 3D print the top and bottom of the frame/cylinder flat, and then cut out pieces of this mylar, and stick them on the flat surfaces.
Once the sand sets, pull out the mylar (it is very slick/slippery), and that should be enough clearance to pull the pattern from the mold.

If that does not work, I will go with a retracts.

I noticed on the bearing cap I cast (photo below) that the bound sand will pick up very fine detail, and the blue painter's tape on the pattern looked like a large protrusion up from the casting surface.
I would guess a good bound sand mold with ceramic mold coat could rival the surface finish of Petrobond, but in iron.
The bound sand on the bearing cap picked up the wrinkles in the surface of the blue painter's tape.
And one certainly does not want to use the old-style pour-basin, sprue, runners, gates, etc. methods (in my opinion), since they tend to cause a lot of internal casting defects, such as bubbles, slag inclusion, entrained sand, turbulence defects, etc.

For surfaces that are not going to be machined, the draft angle does not matter.

I am going to try the mylar on the flywheel too, and the flywheel rim will be flat on the sides and top.

Many folks seem to make castings the way they were always made, because that is the way things were always done.
I think there is much to be improved upon from the old casting methods, if one thinks outside the box (outside the flask, LOL).
I often find that it is best not to follow the various books, old and new, about foundry/casting work, since they seem to universally demonstrate old obsolete methods and materials.

So the plan is no draft angle on the machined surfaces, and perhaps just a few thousanths extra material that has to be machined off to get a flat.
With today's 3D printers, and bound sand, there is really no excuse for the distorted castings that I often see in some engine kits.

.

 

[GreenTwin]

In Solidworks, you can assemble the engine parts, and actually run an engine virtually.
If the parts don't fit correctly, then the engine will not run in simulation.

This is a rough quick mock-up, with a temporary connecting rod, just to try out the simulation.

The entire engine at some point will be assembled, including cam, gears, valves, etc., and then the entire engine can run in simulation, and the valve events observed and corrected as required.

This is a very powerful design feature for sure.

When I built the green twin, I ran it in simulation, and it ran perfectly after a few tweaks.
I told the Live Steam editor that the engine would run, but he did not believe me, and insisted that I built a physical engine and run it before starting the magazine article.
The simulator tells no lies.
If it runs well in simulation, chances are very good that it will run well in real life.

One also has to check for inteferences, which I have not yet checked.
The bump-outs on the sides of the crankcase on the 4hp ball hopper monitor seem like a bit of an afterthought, and looking at the simulation, they don't really seem to be in the right place, although that is where and how they are located on a full sized 4hp engine.

If you double-click on the video below, it will display full screen.
Hit "ESC" to get out of full screen mode.

These days, using modern tools, there does not have to be any significant error in an engine design, and perhaps no errors at all.

 

[GreenTwin]

This is sort of what the water jacket and bore cores will look like.

I will have to thicken the core that is around the entry port to the cylinder, since it gets too thin.

The top of the water jacket core will have to extend upwards into core prints.

The engine could be cast without the water jacket, but I will attempt to cast it with the jacket.

It may be best to just leave that core part over the intake port out, and cast that area solid, since that thin of a core will have a high probability of failure.
I don't think that small area really needs water around it anyway.

.

 

[GreenTwin]

And here is a rough approximation of the corebox for the frame.

Me and the boolean/subtraction thing are not getting along well at all.

Boolean is something I seldom use.

There will be core prints out the top of the bore, out the bottom of the frame, and out the access door opening.

The interior of the 3D printed frame will be used as a corebox, with extensions to create the coreprint areas.


.

 

[Jasonb]

Many folks seem to make castings the way they were always made, because that is the way things were always done.
I think there is much to be improved upon from the old casting methods, if one thinks outside the box (outside the flask, LOL).
I often find that it is best not to follow the various books, old and new, about foundry/casting work, since they seem to universally demonstrate old obsolete methods and materials.

I think you need to differentiate between home casting and commercial casting. While it may be fine for you at home to play about with pieces of mylar which will take time. It is not really practical to do the same with patterns that will be sent to a foundry for possibly multiple parts to be cast. The extra time will cost and the chances of some bits of Mylar going missing are high. Note time is not just removing the Mylar but the fact that if the sand starts to break away then it either needs patching up or moulding again. For the same reason separate core boxes are better than trying to use hollow patterns to do both.

I'm not sure the Near net is always ideal, thin parts are at risk of chill so will be hard to file to a finish, adding machining allowance including draft can be used to advantage to thicken areas at risk of chill such as the corners of thin feet or the tops of valve guide bosses by allowing a little more metal you 1. reduce the risk and 2. allow enough so any chilled metal can be removed with carbide tooling so you net surface is nice soft iron.

So if your patterns are to be made available for posterity then think about how other may have to produce castings from them.

 

[GreenTwin]

The draft angles could easily be toggled on and off in the 3D model, which is how I normally do machining allowances and draft angle in places such as the outerside of the flywheel rims.

Not really a problem to print patterns with and without.

Thin parts are not really at risk of chill if you get the ferrosilicon level correct.
There is a fine line on the ferrosilon level, and it has to be pretty exact to avoid excessive shrinkage and hot tears.
I have had gray iron spills and overruns that were quite thin, and still no chill.

Gray iron poured without ferrosilicon will have thin sections that are the hardness of tool steel, but that can easily be avoided.

People assume thin parts will have chill, but it is not necessarily so.

I am contemplating stacking sets of 5 molds each for each part, or perhaps flat molds with radial pieces of the same part.

With resin-bound molds, you could probably stack five (perhaps more) molds on top of each other, and cement a runner on the side, with manually made gates.

Once I get all the parts 3D modeled, I will probably print out a set of preliminary patterns, and play around with some layouts.

I have cemented molds together in the past, and that method works well to make multiple castings from a single pattern.

I would still have to set up drawings, but I have a lot laid out in 2D CAD, and can import 2D from Solidworks that is derived from the 3D parts.

This is a long term build/project, and so nothing is going to happen very quickly.
.

 

[GreenTwin]

One trend in the casting industry is called the "Additive" process, which is sand molds which are made using large 3D printers.

This process is expensive, but has a very fast design-to-casting-to-machined-part time that can be just a few days.

For folks in the US who are operating large heavy equipment, the supply chain has become to long, and too fragmented.
The operators lose a fortune for every hour of downtime on critical equipment, and so being able to get a replacement casting quickly is all imporant, and the cost of the casting and additive process is secondary.
Wait time for overseas cast parts can be months.

The benefits of the 3D printed molds are that changes can be made at any time to the 3D model, without having to retool any equipment.
3D printed molds are also very popular with the rapid prototyping process, where many test castings must be made and tested in a short period of time.

There are no tooling costs with 3D printed molds.
Tooling costs for permanent molds is very significant, as is the cost to re-tool or modify existing tooling.

This is one of the modern foundry trends that has intensified since the COVID supply chain problems got so bad.
Lots of reshoring of cast parts too, moving the casting process back to the US to eliminate supply chain and quality control issues.

.

 

[GreenTwin]

I got the frame/cylinder mostly done.

The feet and associated webs were very tricky to model with the draft angle going into the body of the part.

That was time consuming, but will allow the pattern halves to be pulled from the sand.

I added some bosses on the feet, which can be toggled off in 3D modeling for the folks who don't want those.
I would use retracts to make them work, but they would give me a flat surface to bolt to.

.

 

[GreenTwin]

I think I am going to cast a base for the ball hopper monitor to sit on.

Below is the base for the Frisco Standard.
It would be similar to this, but shorter, and perhaps more rectangular so that it would fit between the flywheels.

.

 

[ajoeiam]

In Solidworks, you can assemble the engine parts, and actually run an engine virtually.
If the parts don't fit correctly, then the engine will not run in simulation.
snip

These days, using modern tools, there does not have to be any significant error in an engine design, and perhaps no errors at all.

snip

You bet - - - - there don't have to be any errors at all - - - and yet look at how modern design is fairing.
It makes the companies a LOT of $$$$ but repairability - - I'll be polite and just say - - it generally sucks.
Modern design tools are touted as obviating the need for any practical experience.
In my experience - - - engine design is sorta like clothing design - - - its not about how much stuff you
can put together - - - but how little you actually 'need' to get the job done.

"Elegance is merely simplicity refined to its ultimate consequence."

(Today design is a tool to manipulate the masses - - - far more than even in the halcyon days of the 50s and 60s in car design.)

 

[GreenTwin]

The first engine I drafted in 2D CAD was one of my dad's engines.
I disassembled the entire engine, measured every part meticuluously, and drew every part to an exact size, down to the thousanth's of an inch.

The I l realized that the parts had to fit together rather precisely, and the mating holes had to align, etc.
And I realized that standard shaft and fastener sizes were used, that did not match my measurements.

I saved those drawings as a monument to stupidity, but one has to start somewhere, and mistakes will always be made along the way.

I got better at drawing engine parts in 2D, but I still had problems getting an exact fit between mating parts, and figuring exactly how parts with multiple compound curves actually looked in real life.


I started learning 3D modeling, and quit multiple times over a period of a year, often in total disgust because I could not figure out how it all worked.

I kept at it though, and the green twin design was the first engine were the 3D really came together, and worked correctly.
The green twin engine was originally started as just a learning exercise in 3D modeling, and it turned out so well I decided I had to cast an engine.

I have been working on a large work project, and one of the team companies I am working with decided to do much of their work in 3D modeling, "because it was so accurate". I am not finding their 3D model to be very accurate, and there are some errors in it.

The good thing about 3D modeling an entire complex multi-story industrial buildng is that you can do a virtual walk-through, and see very vessel, pump, pipe, beam, stairway, elevator, control room, beam, footing, etc., and that is an extremely helpful feature when designing dense industrial facilities.

The problem is, the 3D model is only as good as the person inputting the data.
The old adage "garbage in, garbage out" still is very true today.

Many 3D models look very good on the screen, especially the high resolution renderings, but if you want to build a real engine, you have to add draft angle, machining allowances, etc. to the 3D model, and that requires quite a bit more thought and effort.
I leave the high-res rendering to others, and concentrate on making workable 3D-printed patterns.

The end goal for me is to make engine castings.
I see 3D modeling as a virtual tool that can be used to derive patterns; no different than a mill, lathe, CNC machine, etc.

I came from the days BC (before personal computers), and we stood at large drawing boards and hand-drew all of our engineering drawings, generally on 30"x42" sheets of vellum.
Many folks still had their slide rules hanging on the wall, although electronic calculators had taken over by about the mid 70's.

There were lots of errors in the early construction days, such as walking down stairs, and there is a beam 48" above the stair tread midway down the stairway. You would be surprised how often this happened in large construction projects.

Our "computers" were IBM punch cards used with a mainframe; and line printers.
No screens, no mice, no USB ports, no flash drives, no hard drives; just you, the punched cards that contained your FORTRAN program, and the IBM mainframe did the number crunching.
Folks went to the moon using archaic stuff like that, and with slide rules, because the guys who programmed the computers were smart, not because the computers were sophisticated. FORTRAN was and is an extremely powerful program in the hands of a skilled programmer, and that was the star of the show, not the computer or printer.
The mainframe ran glacier slow, although the line printer was ok for its time.

Todays 3D modeled equipment does fit much better, if the guy who is plopping the stuff in place on the computer screen is paying attention.

The achiles heel of computers is people.
Your computer stuff is only as good as someone's ability to accurately use it.

Good inputs can create some fantastic outputs.

I feel that 3D modeling and 3D printed patterns are the 21st Century form of model engineering.

.

 

[GreenTwin]

I will be taking a break from this design for a while, to catch up with some work projects.

I am very pleased with this engine design, and how accurate it represents the photos I have been sent by folks who own full-sized Monitors.

All of the surfaces have draft angle included, and so the patterns will be ready for molding once they are 3D printed and filled.

Just a little more work on various bits and pieces, and the valve chamber, and then I think we will be ready for 3D printing.

I would say rescuing this classic old design has been a success, and it has been a dream of mine for many years to document this engine, cast it, and save it from fading into obscurity.

More to come.

Pat J
.

 

[minh-thanh]

.
View attachment 151427

View attachment 151428View attachment 151429View attachment 151430View attachment 151431View attachment 151432

Hi !
When you finish designing, do you intend to sell the plan?? A beautiful engine. Maybe...one day I'll build this engine but I'll scale it down

 

[GreenTwin]

I have had discussions with a lot of people about what to do with this design.

I am definitely going to send the 3D models to several folks, so that it does not get lost, with the understanding that it is for personal hobby use only, and not commercial use. I don't want this engine design to get lost like so many kits these days.

One individual I have talked to has connections to a foundry that will do lose patterns for model engines, in iron, and so I have considered getting them to make multiple sets of castings. I don't know the details of that foundry, and have not talked to them, so if I go that route, there would be many details to work out.
If I did get some castings made, I would probably go with a 10" flywheel, since most model engine seem to prefer the smaller scaled engines.

I don't really want to get into the model engine kit business, so I am trying to figure out how to get castings made, and distribute some of those without having to start a company of some type.

As far as drawings, I would want to be careful about distributing those.

I guess I would give out plan sets to folks that I would trust would use them for personal builds, and not for commercial purposes.
I would not sell drawing sets.

.

 

[GreenTwin]

Looks like the overall engine with 12" flywheels in gray iron is approaching 96 lbs.

With a cast base (probably an aluminum base), the overall weight will be in the 110-120 lb. range.

I guess I would move the engine around on a 2-wheeler, and install a little forklift mechanism with a winch, to raise and lower it in and out of the car, and up onto the tabletop.

I can deadlift 110 lbs., but that is really not good for the back to do that very many times, and so a dolly/winch would really be necessary.

I would make a box to go around the dolly, and just store the whole thing as a unit.

The overall height of the engine would be 24", measured from the bottom of the flywheel to the top of the hopper cap.
With the engine mounted on a base, the overall height could be 26" (+ -), assuming the base is about 5" tall.

I could cast the engine in aluminum, but I have had a lot of trouble with tool bit loading with aluminum.
Heat treating the aluminum helped a lot, but is an extra step that has to be done very carefully in two steps over about 10 hours.

Gray iron drills, taps and machines so well that I think it is worth dealing with the extra weight.

Edit:
I am going to set up drawings for engines with a 12" and a 10" flywheel.
I will look for some suitably avaible gears for a 10" flywheel engine, and draw around those.

I think most will want to build an engine with a 10" flywheel, not a 12" flywheel.

.

 

[Jasonb]

Don't forget to figure in the weight of things like ignition system, battery, nice pully to go onto the flywheel and all the other small parts and castings you have not yet modeled as they soon add up

If you do go with that cast stand you will still want somewhere for ignition and battery so it may be easier to go with the hand sled which has the advantage of not overly increasing the height and being wider makes the model less top heavy and prone to fall over. Alternative would be a drop cart and then you can just wheel it around.

 

[L98fiero]

You bet - - - - there don't have to be any errors at all - - - and yet look at how modern design is fairing.
It makes the companies a LOT of $$$$ but repairability - - I'll be polite and just say - - it generally sucks.
Modern design tools are touted as obviating the need for any practical experience.

I'd argue the repairability is a secondary concern to ease of manufacture. At one point I thought I could be a car salesman, but I couldn't bring my ethical standards low enough, and in the mean time, learned a bit about the fun of repairing cars. I had worked on cars in the 60s when you could actually see the engine you had to repair but then saw a car where the battery was under the windshield washer fluid tank. Now what kind of wizard could have come up with that? Repairability is way down the list of requirements of design.

 

[GreenTwin]

At one point I thought I could be a car salesman, but I couldn't bring my ethical standards low enough

LOL, my brother went to a dealership to buy a new car the other day, and he wanted to trade in his used electric car.

The salesman gave him a very low offer for his electric car, and my brother said "why so low?".

The salesman said "Your car has a bad transmission".

My brother said "Its an electric car, there is no transmission".

The salesman said "Yes, we can tell with modern diagnostics".

My brother said "What diagnostics are you using ?".

Salesman said "There is a special light that lights up on the dashboard".

My brother said "Show me this light".

Salesman points to the "Parking brake ON" light.

My brother said some unpleasant things to the salesman which I will not repeat, and my brother went to another dealership to buy a car.

.

 

[Richard Hed]

Looks like the overall engine with 12" flywheels in gray iron is approaching 96 lbs.

With a cast base (probably an aluminum base), the overall weight will be in the 110-120 lb. range.

I guess I would move the engine around on a 2-wheeler, and install a little forklift mechanism with a winch, to raise and lower it in and out of the car, and up onto the tabletop.

I can deadlift 110 lbs., but that is really not good for the back to do that very many times, and so a dolly/winch would really be necessary.

I would make a box to go around the dolly, and just store the whole thing as a unit.

The overall height of the engine would be 24", measured from the bottom of the flywheel to the top of the hopper cap.
With the engine mounted on a base, the overall height could be 26" (+ -), assuming the base is about 5" tall.

I could cast the engine in aluminum, but I have had a lot of trouble with tool bit loading with aluminum.
Heat treating the aluminum helped a lot, but is an extra step that has to be done very carefully in two steps over about 10 hours.

Gray iron drills, taps and machines so well that I think it is worth dealing with the extra weight.

Edit:
I am going to set up drawings for engines with a 12" and a 10" flywheel.
I will look for some suitably avaible gears for a 10" flywheel engine, and draw around those.

I think most will want to build an engine with a 10" flywheel, not a 12" flywheel.

.

I like how you intend to box these items, however, it would be better to put them in TWO boxes under 50lbs each as fedex, ups and all those delivery peeps will deliver two 50lb boxes for cheaper than one 100lb box. Reason is that 50lbs,, any person can lift it to put on a dolly, but a 100lb box, there are going to be a lot of insurance claims. I had a problem with some B**turd this summer. I had won a lot in auction. The guy who put it in boxes and sends it out wanted to put it all in one box and send it for 390$!. However, he could have put it in four boxes and sent them all to me for 90$. But he acted like that was an imposition on him. I mean, I was PAYING him to box it and send it. How about if HE tried to accommodate his customers? I would have gladly done this for him--lower his total costs, but he didn't care about other poeople's costs! It was no skin off his b@lls, what did HE care? Well, the companyt I was dealing with cancelled my order because he would not send the stuff in a reasonable and timely manner.

 

[GreenTwin]

I worked a bit more on the Ball Hopper Monitor 3D model.

I created the valve chamber, which was more complex than I anticipated, with some odd things like a curved front.

I reworked the flywheel from the test flywheel I made years ago, and got this one much closer to a full sized 4hp flywheel.

And I mocked up a quick base, just to see what the engine would look like on a cast base.
The base is by no means final, but I think I like the looks of it so far, and will refine it a bit more.

.