30cc Inline Twin 4-stroke Engine based on Wetsbury's Wallaby

Help Support HMEM:

Eccentric

Well-Known Member
HMEM Supporter
Joined
Dec 29, 2020
Messages
68
Reaction score
88
Location
Southern California
My next engine build will be a derivative of Edgar Westbury's Wallaby, first designed for a model train, then updated for use in a model hydroplane. The engine is an overhead valve, water cooled, 30cc, inline twin cylinder 4-stroke. My version will use no castings.

I am a huge fan of Edgar Westbury's work as a model engine designer. He was very prolific producing designs optimized for construction by the home machinist with minimal tools, typically a small lathe and a drilling machine. His first engine design was published in the 1920's, a single cylinder 2-stroke for a 13 foot wing span model airplane. He joined Model Engineer magazine in the early 1930's and went on to publish many wonderful engine construction articles as well as several books on machining.

This is a video of a classic Wallaby:

Construction articles for the Wallaby are available from the "Model Engineer" magazine and have been in the public domain for decades. Anyone needing help locating them, feel free to PM me.

The engine has a bore of 1 inch and a stroke of 1-1/8 inches. It has a built in oil pump to provide pressure fed lubrication to the tappets, crank shaft center bush and the connecting rod big ends. I have redesigned the Wallaby to be machined from raw stock, using no castings, and I will be using ball bearings on the crankshaft, camshaft and timing gears.

I use SolidWorks for my computer aided design work and Fusion360 for tool path generation. I first fabricated 3D printed models before committing myself to cutting metal.

The original Wallaby had 5 main castings: the Body Casting, Sump, Cylinder Head, Cylinder Head Plate and Timing Cover.

I have split the Body Casting into three machined parts, the crankcase, rear timing plate and the block. I learned the technique of separating the crankcase and the block from Terry Mayhugh's build of Ron Colona's Offy. When designing a casting, a good engineer will strive to incorporate as many features as possible. This will reduce total part count and reduce cost of a mass production, but results in a very complex part, making it difficult to machine a one off. The Cylinder head has many internal passages that make it a good candidate for a casting, but difficult to machine, so it will likewise be made as two parts to be bonded later with high temperature structural adhesive.

Other design changes include the use of bearing housings for the crankshaft end bearings and the camshaft bearings. This will allow me to fabricate these precision components on the lathe independently from the machining of the crankcase.

-------------------------------------------------------------------------

My first Goal - Machining the Large components:

  • Sump
  • Crankcase
  • Block
  • Timing Back Plate
  • Crankshaft Main Bearing Housings
  • Dummy Crankshaft
  • Camshaft Bearing Housings
  • Dummy Camshaft
1621641528965.png

Assembly model of only the large components


1621641598542.png

Another View


1621641626932.png

3D Printed Mockup of the engine


1621641662370.png


1621641679802.png


1621641693988.png


Sump

I have decided to machine the sump first. It is the second largest part and has many of the same shapes that will need to be machined into the crankcase, so will provide good practice developing the tool paths, tool selection as well as the speeds and feeds. The sump is less critical dimensionally than the crankcase. The crankshaft center bearing is held exclusively in the crankcase and the sump only holds the two bearing housings for the crank case, but will rely on the crankcase for alignment. There will be two locating pins where the sump and the crankcase mate to provide positive, repeatable alignment between the two and the crankshaft bearing housing mating surfaces will be machined from the same setup to provide the best alignment.

Datums for the sump, Top, Right side and Front

Machining Steps

  • Flatten top and Right side of the work piece and insure they are perpendicular
  • Mount the work piece in the vise with Top against the vise face and the Right side down against parallels - machine left side at least .05" over dimension.
  • Mount work piece in vise with the Top against the vise face and the Front down and machine the back side at least .05" over dimension.
  • Mount in the vise with the right side against the vise face and the top down. Machine the bottom square at least .05" over dimension.
  • Take to drill press and drill a 1/2" hole from top, center of sump measured from Right side +.02", from Front +.02", and 1.5" deep (which is more than 1/8" from inside bottom of sump). this hole is the starting point for the end mill to machine the inside.
  • Mount in the vise with the Right side against the vise face and the top up.
  • Machine the inside of the sump. See detail below.
  • Center drill all holes
  • Drill all holes to depth including the two alignment holes. The alignment holes reamed to 1/8" interference fit.
  • Flip part over with Top against parallels, and the Right side against the vise face.
  • Machine minimum clearance for crankcase bolts, leaving the majority of the material to be removed later.
  • Lightly sand the top surface of the sump on a flat surface with 180 grit, then 320 grit and finally 600 grit to create a flat clean surface-don't get carried away, enough to just remove tooling marks.
After these steps, the sump is compete for now and is ready to be attached to the crankcase for the machining of the front and back. 3D modeling programs allow "configurations" of models to be created that are derivative or different than the base part. I use this feature to create modified versions of the model to optimize the machining, creating both models specifically modeled for a specific machining operations as well as special configurations of the stock material. This is nice because as design changes are made to the base part, they are carried forward into the derived parts.

Detail of sump machining operations - A special 3D model is created for the machining of the top of the sump, it has the following modifications:

  • All features are removed from the Front and Back.
  • .020" of material has been added to Front and Right side.
  • Corner fillets removed.
  • Oil drain hole in bottom removed.
A special model is created for the raw stock, it has the following characteristics:

  • square block, .020" over sized on the Front, Back, Left and Right sides.
  • Has a 1/2" hole 1.5" deep in the middle of the sump for the 1/4" end mill that will be used to remove the majority of the material to start in. End mills are great at cutting on their side, but not so good at machining down. By machining from the side of the part, or from a predrilled hole on an interior feature, the pocket machining operation can remove more material, quicker, with less stress on the cutter.
1621641737373.png


Work piece locked in the vise ready to machine the inside of the sump. The 1/2" hole is used for an entry point for the end mill.

1621641760714.png


Modified model for machining the inside of the Sump. Compare to sump model above.

1621641789123.png

Finished inside machining of the sump

1621641817426.png


Finished bottom machining of the Sump. The sump can now be bolted to the crankcase for further machining.




Lessons learned:

  • I added .020" of stock to the top of the model that was machined off during the first horizontal milling operation. I then switched tools and touched off on the top of the model and my tool path was then .020" too deep. I noticed quickly, but there is an area on the back where the main bearing mount is .020" too deep.
  • When machining the bottom I had problems with the work piece remaining securely clamped in the vise. It was moving during machining. I need to experiment, it may of been due to the fact that the work piece was hollow between the vise jaws, it might have been I just didn't have the vise tight enough, it might have been I did not have a three point clamping points, or I did not have enough of the work piece in the vise.
  • Should have used a smaller step over when using the 1/4" ball end mill on the main bearing housing surfaces.
Things that worked well:

  • I used a two flute 1/4" flat end mill at 8,000 RPM spindle speed, coolant, 15 ipm and .050" depth of cut. I was happy with the rate of material removal and the finish.
  • I did not use a ball end mill on the inside of the sump. My thinking was no one would see the inside, so smooth surfaces were not worth the machining time. However, the surface turned out acceptably well with just with the 1/4" flat end mill.
  • When cutting deeper than the cutter flute length (.75" in my case), I step the material out .010" in the model to prevent the sides of the tool rubbing. This effect can be seen in the deep channels cut in the sump bottom.
 

Mechanicboy

Well-Known Member
Joined
Jul 10, 2013
Messages
760
Reaction score
252
Hi, you can cast the engine block and other engine parts when you are use the 3D printed pattern in the cast sand and pour direct into, then the pla will burn away. It save a much time.
 
  • Like
Reactions: gld

Jasonb

Project of the Month Winner!!!
Project of the Month Winner
Joined
Mar 30, 2008
Messages
2,577
Reaction score
528
Location
Surrey, UK
Also look at what notch you had the vice cross bar in as you can end up with more force going downwards rather than sideways if it is too close to the fixed jaw. I tend to only use my one of these vices where I need the height or jaw opening as I've had stuff move the rest of the time I use a Kurt Copy on the CNC as I find it gives a better grip.

An Adaptive type tool path will give more even wear along the side of the cutter when clearing out the waste in parts like the sump, spiral down say 0.2" each stepdown and then cut using 10-15% of the cutter diameter, set a fine stepdown of of say 0.05" to get the final stepping of the contour.

For something like the bearing housings I would leave metal on them and line bor the two halves when bolted together.
 

Eccentric

Well-Known Member
HMEM Supporter
Joined
Dec 29, 2020
Messages
68
Reaction score
88
Location
Southern California
Jason,

Ahh, I see. that makes sense what you say about the vice cross bar and the direction of force depending what notch it is in. I had not thought of that before. I have limited Z and a Kurt style vise is too tall for my machine.

I will try your suggestion regarding the use of more of the side of the cutter (.2" vs .05") with smaller % of the cutter used each pass. thanks.

I agree with you and will line bore the crankshaft bearing surfaces, I have not done this before so it will be yet another adventure. Mounting the engine perfectly in line with the spindle on the cross slide seems daunting.

Mechanicboy,

I like you casting technique. Casting is something I would like to learn how to do at some point. Right now I still have alot to learn about machining. thanks for the post.
 

Eccentric

Well-Known Member
HMEM Supporter
Joined
Dec 29, 2020
Messages
68
Reaction score
88
Location
Southern California
Working on the top crankcase. Here is my plan:
  • Square the stock up on all sides in the lathe.
  • Drill the camshaft hole through the work piece, working from both the front and back, .05" undersized. Use the bottom and right side as the datum.
  • Mount in the mill and using the bottom and the right side as datum, drill and ream the camshaft bearing holder holes from each end.
  • Drill a 1/2" hole approximately in the center of the cylinder as a clearance for the flat end mill to enter.
  • Mount in the vise with the bottom up and the right side against the vise face
  • Machine the bottom using a 1/4" flat end mill. Touch up the crankshaft bearing holder surfaces with a 1/4" ball end mill. Used a .200" step down ( .050" fine step down) with a .050" side cut. 1 hour, 16 minutes total machining time.
  • Machine three flats on the center bearing holder surface, one for a center oil hole and the other two for the middle camshaft bearing mounting screws.
  • Mount in the vise with the top up and the right side against the vise face and machine the top with a 1/4" flat end mill. 28 minutes machining time.
  • Center drill, drill and tap the 6 block mounting holes.
1622241454190.png

Squaring up the stock in the lathe. I left it over sized, but did not need to "re-square" up the work piece after the camshaft hole was drill and reamed.

1622241553146.png

Finished bottom machining. You can see small steps .0745" down. the wall sides are .010" thicker so the 1/4" end mill can mill deeper than its .750" flute length. this keeps the mill shaft from rubbing.

1622241680638.png

Machining the crankcase top surface

1622241736320.png


I also worked on the block, I brought it to exact dimension on the lathe before machining out the inside.

1622241875732.png


1622241915864.png


The engine so far.
 

Eccentric

Well-Known Member
HMEM Supporter
Joined
Dec 29, 2020
Messages
68
Reaction score
88
Location
Southern California
The cylinder head is one of the more complex parts due to the fact that there are 7 different surfaces that need machining, five on the outside and two internal, and they must all align as well as possible. The tools that are used include a 1/4" end mill, 1/8" end mill, 1/4" ball end mill, 3/32" ball end mill, spot drill and drills. I did some significant redesign of the head making the following changes:

  • Used larger, yet more cost effective, CM-6 sparkplugs. I had to juggle their position to clear the head mounting screws.
  • I like to use complete bronze valve cages which incorporate the valve guide and the valve seat. the original design has the valves seat directly into the aluminum of the cylinder head.
  • Moved the water jacket holes to give as much clearance to the cylinder head holes, that is maximize the amount of head gasket material between the holes, the edge and the combustion chamber.
  • Rotated the exhaust flanges so the mounting holes do not hit the seam of the top and bottom halves. I don't want the flange mounting screws putting a separating force on the two halves.
  • Adjusted the sparkplug depth and angle to give good access to the combustion chamber, but not interfere with the valve guides.
  • Maximize the water jacket volume without compromising wall thickness.
In order to machine the internal passageways for the air/fuel mixture and the exhaust gases, I decided to fabricate the head from two pieces bonded together.

1622572480565.png

3D Model of the one half of the cylinder head showing internal passageways

My plan evolved as follows:

  • square up two blocks of aluminum, .25" oversize from front to back and from left to right. Exact dimension top to bottom. Total height of the head is .875, bottom half is .475" thick and the top is .400" thick.
  • Machine internal passages. roughing will be done with a 1/8" end mill entering from the outside of the work piece leaving .010" material. All radii are .1875", so a 1/4" ball end mill will be used for the final finishing passes. Total machining time is 34 minutes per half.
  • install locating pins located in the excess material
  • bead blast the internal passages and the mating surfaces, this increases the surface area for the adhesive.
  • Bond the top and bottom halves of the head using structural adhesive and a sprinkling of 80 grit glass beads. this insures there is a micro space between the halves and the adhesive does not get squeezed out during clamping. Loctite EA9340 is used as the structural adhesive. It has excellent resistance to chemicals including fuels and coolant and is rated to a very high temperature. The alignment pins are used to insure proper alignment, but the three .375" internal passageways can also be used to align the two parts by installing matching dowels. Note: Using the glass beads did not work as they were too big and the two parts just slid around on them like ball bearings and the adhesive would have to be much thicker than I wanted. I ended up just bonding clean parts, but was careful to moderate the clamping force so as to not squeeze out the adhesive.
  • After the adhesive is cured, square the parts up to proper dimension all around.
1622572522342.png

The two starting work pieces next to a 3D Printed model of the head

1622572574538.png

Roughing out the internal passageways with 1/8" end mill


1622572618641.png

Internal passage ways after finish machining with 1/4" ball end mill.


1622572652833.png

Cured cylinder head work piece, machined to proper size all around.

1622572682217.png

Machining of the internal water jacket from the top.


The cooling and mounting holes are machined from the bottom as this is the critical mounting surface with the block. I used a 1/4" and 1/8" flat end mills to machine the top. I did not worry about the aesthetics as it will be enclosed and the roughness will not have a material effect of the coolant flow. This is in contrast to the fine finish machining of the air/fuel and exhaust passages that need smooth air flow and thus justified the additional machining time.

1622572759300.png

Machining the underside - combustion chamber, valve guide holes and water jacket holes. I used a 1/4" roughing end mill, a 1/8" finishing flat end mill and a 3/32 ball end mill for the sparkplug hole.

1622572801616.png

Finished machining the bottom with the head mounting holes complete.


The mounting holes were spot drilled and drilled through using peck drilling. this is where the drill bit enters about a diameter of the drill bit then retracts and clears the chips, working its way "peaking" through.

1622572840234.png

I was very happy with the alignment of the internal passages and the valve guides.

1622572880279.png

Simple jig to provide the proper angle for the machining of the spark plug holes.


A 3D model was created for the machining of these holes using the top edge, closest edge as datums. I am not sure about the proper use of the word "datums" in this context, data is plural for datum, any way I used these edges as my zero points for machining.
 

Eccentric

Well-Known Member
HMEM Supporter
Joined
Dec 29, 2020
Messages
68
Reaction score
88
Location
Southern California
1622573022734.png

Machining is complete. I was happy how the external and internal sparkplug holes met. I did not want to machine the spark plug hole all the way through from the outside as it gets very close to the valve guide, so I machined the internal spark plug hole with the same set up as the valve guides.


1622573050968.png

Finished Head after bead blasting. I masked the bottom surface, not sure why, probably doesn't matter one way or the other.


1622573081315.png

Complete head, all holes drilled and tapped.
 

Jones

Well-Known Member
Joined
Jul 9, 2007
Messages
50
Reaction score
3
Great idea bonding the cylinder head together from 2 parts!

Roy Amsbury's V8 used a built up cylinder head as well, though it's silver soldered brass, so it has been done before successfully :)
 

kuhncw

Well-Known Member
HMEM Supporter
Joined
Apr 12, 2008
Messages
803
Reaction score
120
Location
USA ILLINOIS
Very nice work and a good detailed explanation of your process. Will the plate that seals the top of the head be removable or permanently attached.

Chuck
 

Eccentric

Well-Known Member
HMEM Supporter
Joined
Dec 29, 2020
Messages
68
Reaction score
88
Location
Southern California
Chuck,

The Cylinder head cover plate will not be permanently attached, but held down by the same studs that secure the the cylinder head to the block. I am working on the head cover plate now.
 

Eccentric

Well-Known Member
HMEM Supporter
Joined
Dec 29, 2020
Messages
68
Reaction score
88
Location
Southern California
I am trying a new clamping technique I have seen Terry Mayhugh use to machine flat backed parts that require machining all around the outside boarder- temporarily bonding the work piece to a chunk of MDF. My usual way of machining a part like this is to clamp the work piece to the table with some sacrificial material behind it and clamp beyond the machining boundary. I leave tabs connecting the final piece to the work piece so that the part is held during all machining operations. The problem with this is that these tabs require post machining operations to remove them. I usually cut them off with the band saw, then hand file the last remnants of the rough sawn tab. I am not very good at this and usually file some of the surrounding area or do not get a good blend where the tab was. I figured the Cylinder head cover plate would be a good part to experiment with-it is about a 1/4" thick and 1.75" X 3.25".



I first used the fly cutter to get a nice flat surface on the back side of the part, then used 5 minute epoxy to bond the flat side to a block of MDF. I made the MDF block narrower than the work piece so I could use some shims to space the work piece up from the vise jaws and use them to establish my Z axis zero point. I did not clamp the work piece to the MDF, but instead laid a thin film of epoxy on both the MDF and the work piece, them lightly pressed them together to remove all air. Clamping has a tendency to squeeze out the epoxy and I am not interested in maintaining a dimensionally accurate bond as I will be spacing the work piece off the vise, not the MDF.

1623196346087.png

Using a fly cutter to prepare a flat rear surface

1623196386736.png

5 minute epoxy to secure MDF to work piece


I am concerned about the use of coolant while machining as MDF acts like a sponge, soaking up water, swelling and losing all dimensional stability. I fear that using even a light misting may cause problems as the machining operations will be about 45 minutes. On the other hand I don't like running roughing operations where I am removing a far amount of material quickly without coolant; the cutter will load up as the temp of the work piece rises. In the end I rubbed the MDF down with light machine oil and used a small mist of coolant. Let us see what happens.

1623196823043.png

Work piece clamped in the vise with parallels used to provide proper spacing from the vise

OK, lessons learned. Using both sides of the vise to level a work piece does not work. In my case the clamping jaw is taller than the stationary jaw and the part was machining uneven. I noticed this early so I switched to using parallels under the MDF as is traditionally done. However, due to my caviler bonding process, that is using no clamps, the bottom of the MDF is not representative of the plane of the back of the work piece. Also when I switched to parallels, the work piece was lowered and I did not reset my Z axis zero, so the through holes did not go quite through the part. this can be seen in the last picture below.

Oiling the MDF was not sufficient to prevent water damage, As seen below the outside of the MDF swelled .055".

1623196885557.png

MDF absorbed moisture and swelled even though only a light mist was used for cooling



Evidence of the swelling of the MDF can been seen in the final part. The final machining operation used a 1/16" flat end mill and the part rose with respect to the cutter by a total of .023" between the commencement of the machining and the end.

1623196929034.png

.023" trough due to the MDF swelling and the work piece rising during the machining process.

So where do I go from here? I do like the idea of using MDF as a machining substrate as it is much cheaper than using a piece of sacrificial aluminum, for example. The part I attempted the technique with was relatively small and a little moisture on the MDF and the resultant swelling had an outsized impact on the final result. Do I attempt to seal the MDF somehow? Paint? Is it becoming more trouble than the effort saved? I have further experimentation to do.
 

kuhncw

Well-Known Member
HMEM Supporter
Joined
Apr 12, 2008
Messages
803
Reaction score
120
Location
USA ILLINOIS
Quote: "The Cylinder head cover plate will not be permanently attached, but held down by the same studs that secure the the cylinder head to the block. I am working on the head cover plate now. "

Thanks.

Chuck
 

Latest posts

Top