270 Offy

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Terry,
While I agree that there is a certain amount of thermosyphoning of the water and therefore while running an engine the radiator does tend to heat up (model T cooling system) the introduction of a pump should assist in the cooling effort. Not having a CNC machine I machine my impellers with straight vanes but offset from the centerline to give the desired flow across the vanes. My pump body has the outlet tangential to the inside bore of the pump. I have made the impeller shaft bore two different ways, one on center with the housing bore and one with an offset center to form what has been referred to as a cutwater. I have never tested the flow rate of these pumps but even after just a few seconds of running an engine I can feel heat circulating through the radiator.
Your impeller design should work great but I agree with the other posters that you need to make your discharge port tangential to the pump bore.
Attached is the drawing for the pump on my 4 cylinder engine. This design has somewhat of a cutwater that is simpler to machine without CNC. The pump draws water from the block then pumps it to the radiator.
gbritnell
Un able to open DWG file I would like to connect via Email
Mike
 
Hi Mike,
Send me a PM for my email address.
gbritnell
 
Beautiful work. Just being stupid here, almost all the rods I have seen,
or made, are of the basic tapered I beam construction not round. I know you can make whatever you desire. Were the original Offy rods round? roundish?
Just wondering. Thanks for all your input here.
 
My rods were made to Ron's drawing with only a few minor changes. I've never actually seen an Offenhauser rod, but I suspect they were probably designed for clearance inside the close-fitting crankcase. Here is a photo I found of one being offered for sale on ebay:
C23DCD10-21C2-438F-8586-BDAB8E5B47AB.jpeg
 
A simple fixture was made up to support the rod caps while their rear ends were finish machined. All went well, but then I discovered the diameters of the previously machined big end bores had all come out different. After retracing my steps I found the bores had been growing by a thousandth or so with each succeeding operation. My first suspicion was built-up edge on the tool, but the surface finishes were great, and the cutting edge looked clean under a microscope. I made another test cut and it too was oversize. Since the grub screws in both the dovetail and toolholder seemed snug, I have some more detective work to do. Although the Offy's rod journals ended up within a thousandth of their target diameter, these bore variations will result in a unique rod on each journal after all.

The bearings were machined from a 932 bronze. After turning an appropriate blank, each was split using a .008" slitting saw so both halves could be used. A zero thickness kerf would produce perfectly round bearings for truly round journals, but my rod journals are out of round by a thousandth, and so I went for a next best solution.

The blank's i.d. was calculated by dividing the sum of the journal's circumference (including a running clearance) plus twice the saw's kerf thickness by pi. The blank's o.d. was then calculated by dividing the sum of the circumference of the rod's bore plus twice the saw kerf by pi.

I've included a scale drawing comparing one of my bearings with a perfectly circular equivalent. There's a .002" difference between the two which for all practical purposes vanishes when the bearing shells are deformed by snapping them into place in the rod halves. With the bearing installed and the rod tightened around a gage pin with zero running clearance, I could see no light passing between the two. The fits are probably as good as anything I could have achieved using more conventional soldered or wasted half fabrication techniques.

The rod bearings were fitted to the crankshaft with a .0015" running clearance. Although I was prepared for a slight bind as an out-of round journal rotated inside one of these compromise bearings, the fits were all uniformly smooth for their entire revolution.

The rods were finished up by drilling an oil passage connecting the bearings on the big and small ends. A fixture designed to reduce risk to the rod was used for this final machining step. A drop of Loctite 290 (wicking grade thread locker) was added behind each cap-side bearing in order to reduce the chances of a spun bearing later on. The Loctite was allowed to cure overnight with the finished rods tightened snugly around gage pins. - Terry

605.JPG
606.JPG
607.JPG
608.JPG
609.JPG
610.JPG
611.JPG
612.JPG
613.JPG
614.JPG
 
A simple fixture was made up to support the rod caps while their rear ends were finish machined. All went well, but then I discovered the diameters of the previously machined big end bores had all come out different. After retracing my steps I found the bores had been growing by a thousandth or so with each succeeding operation. My first suspicion was built-up edge on the tool, but the surface finishes were great, and the cutting edge looked clean under a microscope. I made another test cut and it too was oversize. Since the grub screws in both the dovetail and toolholder seemed snug, I have some more detective work to do. Although the Offy's rod journals ended up within a thousandth of their target diameter, these bore variations will result in a unique rod on each journal after all.

The bearings were machined from a 932 bronze. After turning an appropriate blank, each was split using a .008" slitting saw so both halves could be used. A zero thickness kerf would produce perfectly round bearings for truly round journals, but my rod journals are out of round by a thousandth, and so I went for a next best solution.

The blank's i.d. was calculated by dividing the sum of the journal's circumference (including a running clearance) plus twice the saw's kerf thickness by pi. The blank's o.d. was then calculated by dividing the sum of the circumference of the rod's bore plus twice the saw kerf by pi.

I've included a scale drawing comparing one of my bearings with a perfectly circular equivalent. There's a .002" difference between the two which for all practical purposes vanishes when the bearing shells are deformed by snapping them into place in the rod halves. With the bearing installed and the rod tightened around a gage pin with zero running clearance, I could see no light passing between the two. The fits are probably as good as anything I could have achieved using more conventional soldered or wasted half fabrication techniques.

The rod bearings were fitted to the crankshaft with a .0015" running clearance. Although I was prepared for a slight bind as an out-of round journal rotated inside one of these compromise bearings, the fits were all uniformly smooth for their entire revolution.

The rods were finished up by drilling an oil passage connecting the bearings on the big and small ends. A fixture designed to reduce risk to the rod was used for this final machining step. A drop of Loctite 290 (wicking grade thread locker) was added behind each cap-side bearing in order to reduce the chances of a spun bearing later on. The Loctite was allowed to cure overnight with the finished rods tightened snugly around gage pins. - Terry

View attachment 120831View attachment 120832View attachment 120833View attachment 120834View attachment 120835View attachment 120836View attachment 120837View attachment 120838View attachment 120839View attachment 120840
Excellent work and engineering explanation. Easy when you know how. Essentially, The same calculations are made by Glacier and Vandervell. I visited both in 1989 to discuss bearing manufacture, with some Japanese experts, who were well impressed. The only extra that you get in a modern engine would be tempted rare metal plating (Indium, etc.) for anti-corrosion. Unlikely the you engine will need that as it is only to extend lifetime by a factor of 5 or more, against corrosion from acids formed in the oils with age and blow-by gasses.
Incidentally, the mass production uses broaching of pairs of shells to get "perfect" repeatable sized parts - at hundreds of pairs every hour. But "pinch" sizing is the same. Mass production sizing and grading is measured using air gauging, to microns! As grinding of cranks, manufacture of journal housings and journals is all 10 times less capable than the design tolerance for perfect fit, all the parts are measured, then graded, and at assembly, the graded bearings are selected to suit the housing and rod diameters. Not something for the one-off engine.
Thanks for the excellent tutorial.
K2
 
Great work!
I'm sure you are properly torquing all the mains, rod ends, ETC., to check fits & clearances, but what specs are using & how were they determined?

John
 
Hi, I'm watching in the background on this beautifully engineered piece of work and admiring the detail you are going to on every part of it.

Looking through the rod, cap and bearing machining you appear to have machined the rods and caps complete as one piece. Did you then split them and re machine the bores? I cant see any photos that show this. Also on the bearings, I see that you did split these but did you only make one true half at a time or do I see that you made one over centre and one under centre piece from each and then mixed these with another set to get two full rounds? Hope I've explained that well enough for you to understand.
Keep up the great work.

Jon
 
Hi, I'm watching in the background on this beautifully engineered piece of work and admiring the detail you are going to on every part of it.

Looking through the rod, cap and bearing machining you appear to have machined the rods and caps complete as one piece. Did you then split them and re machine the bores? I cant see any photos that show this. Also on the bearings, I see that you did split these but did you only make one true half at a time or do I see that you made one over centre and one under centre piece from each and then mixed these with another set to get two full rounds? Hope I've explained that well enough for you to understand.
Keep up the great work.

Jon
Jon,
The rods and caps were machined together at the same time, but there were two separate pieces of aluminum bolted together with the seam between them being the seam between the rod and cap.
The blanks for the bearings were turned a calculated amount oversize and after parting off, the bearing was slit in half and both halves used.

Sorry the write-up was so unclear.

Terry
 
Last edited:
As mentioned at the very beginning of this build, in addition to a slight repositioning of some of the head bolts, the diameters of the liners were reduced by .030" to improve the robustness of the head gasket. These changes in turn required changes to the pistons, and a sketch shows their dimensions and clearances inside the liners. The Offy's pistons use two rings and a groove below them with an array of radially drilled holes for oil control. The pistons were machined from 7075 aluminum.

The wrist pins were turned from drill rod and then hardened. Soft aluminum rivets pressed into their ends prevent scoring on the liners. In order to avoid scuffing and/or abnormal ring wear, it's important for the wrist pin bore to be carefully placed on the piston's centerline and truly perpendicular to it.

After attaching the pistons to the rods, I found a cosmetic filet that I'd added on the rods' small ends was nearly in contact with the inside floor of the piston. I created a fixture to allow me to allow the rods' small ends to be manually rotated against an end mill to remove the fillet and return the rod to its stock .013" clearance.

As far as I can tell, only the rings and the radiator currently stand in the way of the engine's final assembly. Of the two, the radiator with its PITA starter clearance hole through its core seems more interesting and will probably be tackled next. - Terry

615.JPG
616.JPG
617.JPG
618.JPG
619.JPG
620.JPG
621.JPG
622.JPG
623.JPG
624.JPG
 
Beautiful workmanship! WELL DONE sir!
I just wonder about the design? - I would have expected a more "traditional" shape of the con-rod section? I.E. - H-section for max strength and stiffness and minimum weight?
Why the round section? - As you are so capable with the miller, I can't imagine it is because the lathe work is "easier?" than milling the H-section?
I am very impressed by the work and thread.
Thanks!
K2
 
Beautiful workmanship! WELL DONE sir!
I just wonder about the design? - I would have expected a more "traditional" shape of the con-rod section? I.E. - H-section for max strength and stiffness and minimum weight?
Why the round section? - As you are so capable with the miller, I can't imagine it is because the lathe work is "easier?" than milling the H-section?
I am very impressed by the work and thread.
Thanks!
K2
See posts 423 and 424 above ...
 
Thanks Mayhugh1. I had missed that. Picking up threads after hundreds of posts means there are many "answers" already posted that are hard to find. I had guessed it would be for rod stiffness in the available space but it does appear to be a design for a lathe rather than truly optimised.
But your genius s not that you are using someone else's design, but in the way you really understand and explain your approach to such problems as the bronze big-end bearing maching method - and calculations. I am really impressed with your complete approach. I am enjoying learning from you. Well done Sir!
K2
 
I had guessed it would be for rod stiffness in the available space but it does appear to be a design for a lathe rather than truly optimised.
I think you missed the point - look again at post #424 which has a picture of the actual rod from a full-size engine. The rod design is obviously staying as close to true scale as possible. As the full-scale engine was a well refined racing engine, I would say the rod design was optimised for power production, but not necessarily the same parameters as other engines.
 
Thanks Cogsy. I appreciate people trying to accurately copy full sized engines. Beyond my skills and patience. All the workmanship on show on these threads puts my humble work to shame. But I like to know why something has been made the way it is, to learn more of the various designs - and history thereof.
H-section con-rods have been the "optimised" design since late Victorian times, as metals became better controlled, and manufacturing methods allowed parts to become closer to the "Designer's ideal", rather than just "what it is possible to make". Obviously on this racing engine there was some reason to vary from that norm, but the answer isn't here. Possibly drawn aluminium was more "defect free" than forged steel at the time of developing and racing that engine? - or maybe it was a special alloy - tempered for strength - or something? Or perhaps the machined alloy rods were light, stiff enough and minimise stress raisers that start fatigue cracks? - My limited knowledge of racing engines is mostly the British Motorcycle engine developments of the 1940s to 1960s. (I owned a couple of old Triumph racing engines, and I seem to remember the T500C 1955 racing engine having forged aluminium H-section rods....? - But it may have been the 1946 GP engine? - Also, Road Production engines had forged steel rods? - Someone will correct me if I am wrong in my recollection!). The reason for asking about the design is my engine experience of seeing "shapes" used to "maximise strength against minimum material", and these rods deviate from that experience. - Not always easy in the Home workshop. And not intended as any criticism.
Post 431 is "Rocket science" compared to my humble workshop practice. I'm just an Engineer, not a skilled artisan. I am awed by the skill and workmanship applied here.
K2
 
H-section con-rods have been the "optimised" design since late Victorian times, as metals became better controlled, and manufacturing methods allowed parts to become closer to the "Designer's ideal", rather than just "what it is possible to make".

H-section rods abound and have for a long time but there are still significant deviations in some specialised applications (especially high-level racing). Have a look at the profiles used for top-fueler dragsters for instance. Strength and clearance considerations could easily be the reason these rods are non-traditionally shaped, but I think it's unlikely they were chosen simply for ease of construction.
 
H-section rods abound and have for a long time but there are still significant deviations in some specialised applications (especially high-level racing). Have a look at the profiles used for top-fueler dragsters for instance. Strength and clearance considerations could easily be the reason these rods are non-traditionally shaped, but I think it's unlikely they were chosen simply for ease of construction.
It's pure speculation but I'd imagine that the shape of a top fuel con rod has more to do with column strength, they're running around 6000 psi cylinder pressure and at $1500 for a set of eight rods I doubt changing the forging dies to reduce the weight of a 900 gram rod would be much of an issue if a reduced weight could be achieved and the rod still hold up.
As a matter of interest, AWA composites Arthur Warfield & Associates make top fuel rods from an advanced carbon fibre composite, it reduces the weight to about 640 grams and instead of lasting 4 - 6 runs is projected to last 600 runs and as a side benefit when the rod fails it shatters rather than destroying the block. The downside is a set of rods is $144,000.
 
The effectiveness of the model engine coolant systems I've had experience with don't seem to scale very well with engine displacement. There are probably some hard-to-fix reasons for this, but there may also be a few things that can be improved with a re-think of the radiator.

The Offy's block and head, which seem better designed for coolant than most, hold only about 1.5 ounces. This works out to be less than half the amount of coolant per cubic inch cylinder displacement than, say, a small block Ford. The first step in removing waste heat from around the combustion chambers is to transfer it to the coolant, and more coolant can mean more removed heat.

Another limitation involves the relatively poor thermal conductivity of brass which, due to its solderability, is a popular material for model engine radiators. This important material property affects the amount of heat absorbed from the coolant circulating through it - the second step in the engine's cooling process. Thermal conductivities for aluminum and copper, the materials used in full-size radiators, are two to four times greater than that of brass.

A third issue is the typically poor air flow through the core of a scaled-down radiator. Automotive radiators are designed to shed their heat to the air flowing through their core which is the final step in the cooling process. Even a well-shrouded model engine fan has difficulty pulling enough air through the tiny passages of a scaled-down core to be of consequence. The use of an electric fan would be a major improvement, but the core isn't being fully utilized with only conduction and radiation.

The Offy's requirement for a radiator with a starter shaft running through it complicates an already troublesome part of the cooling system. And so, I gave up trying to design a traditional radiator for it and instead focused on increasing the amount of coolant in the system. I used the space that would have otherwise been taken up by an underutilized core to hold it. In the worst-case, the engine's runtime should be lengthened by the time it takes to heat up the additional coolant to an unacceptable temperature. In the best case, the reservoir holding it may perform as well as a traditional radiator core sitting in static air.

As shown in the SolidWorks renderings, the reservoir was designed to look like a radiator with an upper and lower tank attached to a faux core. Since the components are aluminum, they'll be bonded with JB Weld rather than soldered.

As a baseline, the Howell V4 radiator uses traditional finned brass tube construction and has an envelope that's roughly 4" x 6" x 1/2" that holds about an ounce of coolant. The total surface area of its finned core is some 200 square inches.

For comparison, the Offy's reservoir measures 7" x 8" x 1-3/4" and will hold about 40 ounces of coolant. Its all aluminum construction should absorb more heat from the circulating coolant. Even with only radiation and conduction heat transfers, its 150 square inch surface area may shed heat as well as a typical finned core. - Terry

625.JPG
626.JPG
627.JPG
628.JPG
 
Terry,
Could you not use radiators like in you Rolls Royce V12 Merlin?(computer coolant rads), Just a question. I have been following along of another fantastic build
Cheers
Andrew
 

Latest posts

Back
Top