1/3 Scale Ford 289 Hi-Po

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.
Peter, Charles, et. al.

For my last few engines I've been referencing the HMEM discussion thread that began here:

Determining a carburetor throath diameter | Home Model Engine Machinist Forum

D = K * sqrt(c*n)

where D is Venturi diameter in mm
K is a constant .65 to .90
c is cylinder displacement in cc
n is top rpm/2000 (for a 4-cycle engine)

"K is a constant .65 to .90" --- ahhh, there's nothing like an empirical fudge factor :) !!!


wondering how your .180 compares to 1/3 scale ?
 
The revelation to me was looking into the throat of the carb on my Holt engine. At idle there is such a small gap between the throttle barrel and the venturi bore you'd wonder how the engine would run at all. This engine has a 1.00" bore and 1.25" stroke. At that point I make all my carbs small and have never had any performance issues. Yo can always make the bore bigger but not so easy to make smaller
 
"K is a constant .65 to .90" --- ahhh, there's nothing like an empirical fudge factor :) !!!


wondering how your .180 compares to 1/3 scale ?
A 650 cfm Holley 4 bbl (the largest you'd want to put on a 289) has a 1.125" Venturi. If you put 5krpm into the above equation, convert metric to imperial, and use a K of .7 you wind up with 1.02 inches for a needed Venturi. - Terry
 
The included angle of the outlet taper in my last revision was 16 degrees which is within the recommended range that I came across in my reading. It can be easily reduced to bring it within Charles' recommendation, but I'm wondering if these Venturi optimizations really apply to a rotating throttle valve in a carb. The theory seems to have been derived for optimum results at wide open throttle, but I wonder how much still applies at part throttle when the inlet and outlet geometries have changed so radically.
Good point. I would not argue with your 16°. The chapter on 'Turbulent Flow in Pipes' says the disturbance to the flow dowstream of an inefficient fitting like a globe valve may persist for for 50 or more pipe diameters before it sorts itself out. On the other hand, in the case of a straight, smooth bored system without obstacles the optimimum 'diffuser' angle is as little as 6° included.
 
A 650 cfm Holley 4 bbl (the largest you'd want to put on a 289) has a 1.125" Venturi. If you put 5krpm into the above equation, convert metric to imperial, and use a K of .7 you wind up with 1.02 inches for a needed Venturi. - Terry

OK, so 1/3 * 1.125" = .375" is the number I was looking for.
(thou I'm confused by what you said, is that 4 venturis at 1.125" ?)

getting back to scaling laws, the formula you gave has be be pretty much correct because you can re-work it to (throat)^2 ~~ displacement X rpm, which translates to throat area is proportional to air volume (per second), which means all carbs operate best at approximately the same throat air velocity regardless of size, and that's what I was getting at with my dimensional / scaling analysis.

so I'm guessing your throat is so small because you're not planning to operate at 3 x RPM (where scaling laws say you have to be for the same HP per Disp).

I tend towards the same model RPM as (what I'm guessing) you do, we don't want the model engine to "scream" (at super high rpm) but rather just "roar" (at rpms that sound like the full size or not too much more)

Pete.
 
OK, so 1/3 * 1.125" = .375" is the number I was looking for.
(thou I'm confused by what you said, is that 4 venturis at 1.125" ?)

getting back to scaling laws, the formula you gave has be be pretty much correct because you can re-work it to (throat)^2 ~~ displacement X rpm, which translates to throat area is proportional to air volume (per second), which means all carbs operate best at approximately the same throat air velocity regardless of size, and that's what I was getting at with my dimensional / scaling analysis.

so I'm guessing your throat is so small because you're not planning to operate at 3 x RPM (where scaling laws say you have to be for the same HP per Disp).

I tend towards the same model RPM as (what I'm guessing) you do, we don't want the model engine to "scream" (at super high rpm) but rather just "roar" (at rpms that sound like the full size or not too much more)

Pete.
Pete,
The throat diameter doesn't linearly scale. You can't divide the full size Venturi diameter by the linear scaling factor to get the scaled value. The equation shows the diameter scales as the square root of the cylinder displacement. - Terry
 
Pete,
The throat diameter doesn't linearly scale. You can't divide the full size Venturi diameter by the linear scaling factor to get the scaled value. The equation shows the diameter scales as the square root of the cylinder displacement. - Terry

I think you're right, I think I messed up somewhere (wouldn't be the first time :-( !!!)

the reason its to hard to reason about this is Displacement doesn't scale linearly with Bore either, and I was attempting to relate throat bore to cylinder bore, and then you might or might not want to re-scale the RPM also.

in any event I still like the equation, because it implies same air-speed through the throat regardless of size, and I thank you for bringing it to my attention.
 
I think you're right, I think I messed up somewhere (wouldn't be the first time :-( !!!)

the reason its to hard to reason about this is Displacement doesn't scale linearly with Bore either, and I was attempting to relate throat bore to cylinder bore, and then you might or might not want to re-scale the RPM also.

in any event I still like the equation, because it implies same air-speed through the throat regardless of size, and I thank you for bringing it to my attention.
I too am amazed that equation holds over such a wide range. The thread I referred to had several examples of running model engines where the throat diameter seemed to agree with the equation whether or not the equation was actually used or whether those diameters had been empirically determined. The only thing I haven't made my peace with is why the Perry RC carbs I have are so much larger (they're 5/16"). Anyway, I've started making chips on the design I last showed. - Terry
 
Last edited:
The carburetor body and bowls were machined from 6061 aluminum. Construction began with the bowls. Teflon gaskets seal them to the carb body, and o-rings under the heads of a pair of custom mounting bolts seal them to the bowls. A pair of passages inside the body interconnect the bowls to ensure the regulated fuel levels will be the same in both. The bowls are vented to the atmosphere, and a 15 mm watch crystal epoxied in the side of the front bowl provides a view of the fuel level. The rear bowl will eventually contain a fuel inlet and a return standpipe that regulates the fuel level.

The carb body is complicated and has features on all six of its sides that will have to be machined in as many setups. The first step was to square up a block of aluminum. While the workpiece was still easy to hold, all holes normal to its surfaces were drilled and reamed (or tapped).

The integral 'metering block' surfaces were machined next. The rear block contains the already drilled passages associated with the air bleed. The sharply angled main fuel passage through which fuel will be drawn from the bowls was drilled through the front block after it was finish machined.

The carburetor's more interesting top surface with its bell-mouthed inlet were machined next. After flipping the workpiece over, the carb's base and blended outlet were finish machined.

After machining the bowl gaskets and shoulder mounting bolts, the bowls were temporarily assembled for fit checking. Leak checks will have to wait until the rear bowl gets its fuel inlet. All three parts were then bead blasted, cleaned with Dawn dish detergent and alodine'd to create a surface finish similar to the yellow dichromate finishes on the original carbs.

The fuel inlet and standpipe assembly that together will maintain a constant fuel level in the bowls just below the carb's spray bar will be created next. Turbulence created by the fuel being pumped into the tiny available volume usually makes this a trial and error process. The 289's carburetor has nearly twice the volume typically available, although it's split between the bowls. - Terry

546.JPG
547.JPG
548.JPG
549.JPG
550.JPG
551.JPG
552.JPG
553.JPG
554.JPG
555.JPG
 
Crazy nice workmanship. Over and over on this engine !!
Always amazing.

What is your source for the Alodine process chemicals? I love the look. I already can do anodizing. Perhaps it is not necessary to add this process to my repertoire??

Thanks
 
Crazy nice workmanship. Over and over on this engine !!
Always amazing.

What is your source for the Alodine process chemicals? I love the look. I already can do anodizing. Perhaps it is not necessary to add this process to my repertoire??

Thanks
Dave,
It's a one part chemical dip called Bonderite 1201. Henkel has the equivalent, and Amazon carries both. It's commonly used as an aluminum surface prep (before painting) by home airplane builders. If you can do anodizing, it's probably not of interest to you as a final surface as its not as rugged. - Terry
 
The only thing I haven't made my peace with is why the Perry RC carbs I have are so much larger (they're 5/16")

At the risk of deepening the rabbit hole, I tried to find my prior posts showing charts of carb sizes for various model engine commercial offerings, but looks like the links are broke. Anyway, here they are in slightly modified form FWIW. The first 2 charts are overlay of Perry carb bodies (which are predominantly 2-stroke methanol) vs the equation D = (C * n)^0.5 from attached link reference. I (probably mistakenly) call these curves 'Walbro Gas' because at the time I was trying to see the distinction between methanol/gasoline with other factors normalized. But I think the linked article was really more about comparing the sizing equation to Walbro body/engine combinations, so I should probably just re-label them to 'equation'. I think back in the day Perry did make specific bodies for specific engines, but I can't help but thinking they eventually ended up with a progressive size range, so it was easy enough to assign engines to their sizes as pretty close & might explain some stair-stepping here & there.

Anyway, the resultant chart is still a mashup of caveats IMO. Perry carbs, predominantly correspond to 2S methanol single cylinder glow engines. But these engines might vary from 8-12K rpm sport where idling is important to 15-25K race engine where anything below mid-throttle is a non-issue, many just have fuel shut off. Note the outlier Nelson Q40 coordinate, a 0.4 CI with 11mm venturi & even ID is imposed by rules, not necessarily optimized. I believe the link mostly relates to 2-stroke gasoline engines my inferred application, although the article doesn't explicitly say. The equation driven curves displace as a function of RPM. To demonstrate that I'm showing the same chart at 5K & 10K rpm.

Then to complicate matters, a 4-stroke multi-cylinder glow engine venturi is typically lower than its 2-stroke counterpart on cylinder volumetric basis, but we also have overlap depending on cylinder count & timing. And then there is slightly different idling & choke setup. In that regard, the last 2 charts are my comparison of O.S. brand methanol 4-stroke only from published specs. Supposedly some guys keep the same methanol carb for gasoline conversion & other's resize. But that's yet another tangent IMO because with gasoline they are switching to a spark ignition system with advance/retard and/or looking for regulator/pump features for those attributes which only come on gasoline carbs of certain nominal sizes. This subject is a bit of a minefield, at least to me LOL.

https://www.billetboard.com/forum/scooters-gas/scooter-engine-tech-talk/9324-
 

Attachments

  • SNAG-27-04-2024 10.40.56 AM.jpg
    SNAG-27-04-2024 10.40.56 AM.jpg
    55 KB
  • SNAG-27-04-2024 10.41.16 AM.jpg
    SNAG-27-04-2024 10.41.16 AM.jpg
    56 KB
  • SNAGIT.jpg
    SNAGIT.jpg
    53.2 KB
  • SNAG-27-04-2024 10.18.01 AM.jpg
    SNAG-27-04-2024 10.18.01 AM.jpg
    35.4 KB
  • SNAG-27-04-2024 10.18.23 AM.jpg
    SNAG-27-04-2024 10.18.23 AM.jpg
    31.5 KB
Last edited:
Dave,
It's a one part chemical dip called Bonderite 1201. Henkel has the equivalent, and Amazon carries both. It's commonly used as an aluminum surface prep (before painting) by home airplane builders. If you can do anodizing, it's probably not of interest to you as a final surface as its not as rugged. - Terry
Thanks Terry.
 
Thanks Petertha. VERY INTERESTING!
I crunched some numbers for my Moto Guzzi V50II.
Deduced that the factory 24mm carb gave K = ~0.57. (for peak power at 7000rpm).
I considered changing to 26mm carbs (opening the manifold accordingly) to give K = 0.62...
I like the maths - just not sure how applicable to the Del Orto carbs on the bike (always reputed to be too small!).
But THANKS!
K2
 
Inlet and outlet assemblies for the rear bowl were machined from brass. Each is made up of a hose barb and threaded body that was Loctite'd to the bowl. The outlet assembly includes a standpipe which regulates the carb's fuel level. The standpipe is a separate threaded part whose height is adjustable so the fuel level can be fine tuned to just below the carb's spray bar. A tiny additional hole drilled in the outlet just above the bowl's floor allows fuel to drain back into the tank after the pump is shut off.

The bowls were temporarily mounted to the carb body and a trial fuel loop set up using an electric fuel pump circulating white gas. Although designed for RC methanol use, these pumps have been running fine on my last five gasoline engines. Unfortunately they've since been discontinued, but I purchased a number of them while they were available. Occasionally, they'll show up on eBay. The internals of the particular pump in this test will eventually be repackaged for the 289.

The height of the carb's spray bar was scribed on the bezel of the sight glass which shows at best only about half of the bowls' capacity can be used for fuel. The difficulties involved with setting up one of these loops includes dealing with turbulences created by an energetic pump filling a tiny volume. This was mitigated some with a speed controlled motor and a .018" restrictor inserted in the pump's output to reduce the efficiency of the constant displacement pump with a back pressure forced internal leak.

A pesky meniscus (approx. 1/8" high) that forms around the entrance of the standpipe further reduces the usable volume of the bowls. The standpipe adjustment is a trial-and-error process, and although each adjustment required disassembly of the carb, final tweaking required just a couple tries.

The tiny shutter for the air bleed was machined from mild steel and cold blued. A 1-72 button head screw will secure in place after the engine is running and the carb is fine tuned.

All that's left to do on the carb are the throttle and needle assemblies. - Terry

556.JPG
557.JPG
558.JPG
559.JPG
560.JPG
561.JPG
562.JPG
563.JPG
 
The carburetor's throttle which includes the spray bar assembly and rotating throttle barrel were tackled next. First up was the spray bar assembly. In addition to a couple Delrin spacers it consists of a spray bar which screws into the carb body and a needle subassembly that screws into the spray bar.

The spray bar's delicate features and ultra fine threads made it one of the more challenging parts of the carburetor. After starting out with 303 stainless, a switch was made to 12L14 when the threading dies turned their noses up at stainless. Since the features needed to be concentric as possible, all lathe operations were indicated in a four jaw chuck.

A .022" fuel passage drilled through the .040" spray bar nozzle was done using a sensitive drill feed and an accurately aligned tailstock. The 10-56 external threads on one end of the spray bar were simple enough, but the 8-80 external threads on the opposite end were a real headache. The 8-80 threading operation was the last step in the spray bar's machining, and after scrapping three nearly finished parts I started practicing on less precious stock.

An 8-80 thread is only .007" deep, and a clean start with the die in my tailstock die holder seemed impossible even with the help of a couple lathe starting passes. In nearly every case the threads cleaned up only after the entire die had engaged the workpiece. It seemed the clearance stack-up among the die, die holder, and tailstock was just too great for an 80 TPI thread.

Initially the part was made longer and the bad threads cut off, but then cross-threading became a problem. What finally worked was turning a slight starting taper (approximately the width of the die) on the workpiece. After a couple frustrating days I finally had a spray bar I was happy with.

The needle assembly was designed around a common sewing needle, and its 8-80 internal threads went without a hitch. A Delrin bushing keeps the needle centered inside the spray bar as it enters the nozzle, and a spacer protects the whole assembly from damage by limiting the fully closed position of the needle. For best accuracy the needle was JB-Welded in its holder while the whole assembly was temporarily assembled and the needle fully closed.

The throttle barrel was machined from black Delrin for a snug sliding fit inside the carburetor. It's captured in the carb's body with a bolt-on retainer. The head of a judiciously placed SHCS serves both as both an idle and WOT stop. (The idle is adjustable with a screw in the throttle.)

The fuel loop was set up once more. Sanity checks were made on the throttle by blowing air through the carburetor throat and verifying the needle had some semblance of control over the spray pattern below the carburetor. A final tweak was also made to the standpipe height to prevent startup transients from forcing fuel out the bowl vents - Terry

564.JPG
565.JPG
566.JPG
567.JPG
568.JPG
569.JPG
570.JPG
571.JPG
572.JPG
573.JPG
 
The 289's air cleaner is similar to one of the several aftermarket 'Cobra' air cleaners that were popular on this engine. These long open-air cleaners used pleated paper filters encased by metal side screens and rubber end seals. They typically overhung the distributor with some body styles lacking the necessary under-hood clearance to install them. I remember having to use a spacer under mine to clear the plug wires on a '65 Mustang.

Since the little 289's air cleaner had to be designed around its air filter, the filter was tackled first. A round open-air filter scrounged from a stash of old car parts was cannibalized for materials to make the 289's filter. Its rubberized end seals were cut free and discarded after rotating the outside of the donor filter against a slitting saw set up in the drill press. The pleated element and its supporting screens were simultaneously cut through.

After a quick finger count, I continued on pressing my luck. The remaining carcass was wrapped in duct tape and cut down to its final height in the same setup. Although the mesh of the donor's screen looked compatible for the scale involved, the full-size paper pleats made the filter a little too wide. Its width though will be hidden from view inside the air cleaner.

The footprint of the scaled-down filter was machined into a piece of MDF so the filter components be formed and temporarily assembled inside it. A pair of aluminum end plates were machined with shallow troughs that were filled with JB Weld (thinned with acetone) prior to assembly. While still inside the MDF the filter assembly was flipped over, set into one of the end plates, and the epoxy allowed to cure overnight. The second plate was attached similarly. The filter elements were masked off and the end plates painted black for a look and feel of the rubber end seals on the full-size filter.

With the filter in hand, the air cleaner's design could be finalized. It included a baseplate and a decorative top cover with the obligatory 'COBRA' engraved across it. Machined recesses in the air cleaner's top and bottom halves locate and stabilize the air cleaner assembly. The lettering on the top surface was done using 1/8" and 1/16" end mills spinning at 13k rpm using Tormach's spindle Speeder. The machining time for the top cover was just over an hour.

The baseplate was bead blasted and left natural to look like a cast part. The top cover was gun-Kote'd black and the lettered top surfaces sanded with 220g paper. A pair of decorative Delrin nuts completed the air cleaner. - Terry

574.JPG
575.JPG
576.JPG
577.JPG
578.JPG
579.JPG
580.JPG
581.JPG
582.JPG
583.JPG
 
Back
Top