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.
The water neck returns coolant to the radiator and is part of the thermostat housing bolted to the front of the intake manifold. Its small boubous shape and neck exiting the housing at a compound angle makes it a challenging part to machine. Due to an oversight in my manifold design, access to the housing's lower mounting bolt will wind up blocked by the timing cover. With no thermostat to maintain, this isn't a problem for a model, but on a full-size engine this would have left a 60's mechanic screaming bloody murder. On one of today's engines it would be lost in the noise.

Construction began by bolting together a pair of squared-up and dowelled blocks. The thicker of the two was the actual workpiece, while the thinner block was a work-holding fixture that provided machining references and multi-side access to the workpiece. The majority of the machining was done through the front face of the workpiece. Ninety-nine percent of the material hidden behind the neck was accessed by indexing the workpiece 90 degrees around the x-axis. Rather than deal with a third setup, the remainder was removed with a file.

An initial trial fit of the housing uncovered two problems. The first was a minor interference between its bottom surface and the top surface of the timing cover. My SolidWorks assembly had been trying to warn me about this, but I guess I hadn't been listening. Reshaping with a file took care of it.

The second problem was more serious. The housing was to be mounted to the manifold with two SHCS's. The filleted mounting surface for the lower screw (the one that will end up blocked by the timing cover) had to be slightly counterbored so the head of the fastener could sit flat upon it. Unfortunately, an error that I'd allowed to creep into the g-code sent the counterbore completely through the part.
With a half dozen hours already invested, I hoped to not start over. A custom bronze bushing augmented with Loctite was pressed into the bore, and its mounting surface re-machined. The mounting screws were changed to button heads with slightly larger heads, and a custom washer was added for additional insurance against the fastener pulling through. More file work was required to clearance the mounting surface for the top button head.

The coolant passage through the housing required drilling through two faces of the part. The hole through its mounting surface was straightforward, but the neck hole required a sketchy compound setup. An o-ring set into a groove in the intake manifold provides the coolant seal. After bead blasting the housing was Gun Kote'd to match the rest of the engine. A catalyst added to the Gun Kote allowed it to cure at 200F which was a safer temperature for the Loctite used. A stainless steel hose barb machined as a separate item was Loctite'd (using activator) inside the housing. Finally, the housing was assembled to the intake manifold.

With the top of my workbench covered with engine parts, trial fitting new ones has become awkward and a little risky. Years ago while building the Merlin I made a rotisserie engine stand to help with its assembly. Using custom adapters, all my engines since then were assembled on that stand. The 289's adapter supports the block from its bell housing mounting surface. - Terry

Although the alternator will tension the fan belts, it's mostly a cosmetic addition. The goal was to come up with something realistic but substantial enough to be driven by the engine. The final assembly included a half dozen machined parts.

The most interesting piece was the armature shaft's outer assembly which included the alternator's cooling fan and a pair of driven pulleys. The pulleys were machined on the lathe using a cross-section developed earlier for the crank and water pump pulleys. The fan was machined on the Tormach and the surfaces finished with an Alodine dip.

I've been using Alodine 1201 with mixed success to simulate cad plated finishes on some of my engine parts. The surfaces were typically prepped with bead blasting followed by a one minute dip in Aluminum Brite and a thorough rinse in water. An on-going problem has been that after fifteen minutes in Alodine the parts usually took on little or no color. If the dip was extended the parts took on weird blue and green splotches. Others have blamed similar problems on Alodine's relatively short shelf life and mine's coming up on its ten year birthday. However, my recent experiments seem to indicate that my problem may have been related to the bead blasting. The photo of the finished outer shaft assembly shows the results I've been chasing, and it was finally achieved by skipping the bead blasting.

The front half of the alternator's housing was machined in a single setup from a block of aluminum epoxied to a piece of MDF. The front half contains one of the armature's two ball bearings. The housing was bead blasted and left bare to simulate a cast surface.

The rear half of the housing was roughed out on the lathe and then supported in a collet block for finish machining on the Tormach. It holds the second ball bearing. The shaft was machined from drill rod and its outer end threaded for a locking nut to secure the pulley/fan.

The assembled alternator is attached to the engine at two points. The first is a pivot on the front of the right head. A lower bracket locks the bottom of the alternator at the angle needed to tension the belts (#239 o-rings). The lower bracket which was Gun Kote'd black includes an integrated timing pointer located just above the crankshaft damper. The damper is keyed to the crankshaft and will be engraved with timing marks after the valve timing is established. - Terry

The dipstick ...

The dipstick was the last item on my 'loose ends' list. Located in a busy area on the 289, the dipstick tube snakes down the front of the passenger-side head behind the alternator, through the lower corner of the timing cover and into the oil pan.

A realistic looking dipstick isn't difficult to make, but coming up with one that's also functional is difficult. Oil viscosity doesn't scale, and surface tension and capillary action acting on a tiny dipstick tend to wet the whole thing. On top of that, a line made by clean oil is tough to discern on a small dipstick especially after it's been drawn through the tube. If I'm being honest, none of the dipsticks I've made for any of my engines were very useful.

The bends in the 289's dipstick tube exacerbate all these problems. However, I'm trying something different for the 'dipping' part of the 289's dipstick that may help visibility: 1/16" wire rope purchased from the local hardware store. Its flexibility allows the dipstick to easily snake through the tortuous path inside the dipstick tube, and its rough surface tends to hold the oil for a little better visibility. The .006" wire strands making up the rope were silver soldered together at the stick's bottom tip and ground into a ball to facilitate its travel through the tube.

The top part of the stick is a length of steel rod that was turned down and fashioned into a loop for the draw handle using a torch. The top of the tube is clamped to the head, and its bottom is secured inside the timing cover. The top end of the wire rope was silver soldered inside the bottom of the handle, and after bead blasting, all parts were painted to match the block.

The exhaust ...

The original plan for the exhaust was to fabricate a set of stainless steel tubular headers. With our daily temperatures beginning to dip below 100F, I decided to launch another lengthy machining project and create a pair of exhaust manifold 'castings'.

One of the visually distinguishing features of the 289 HiPo was its factory cast iron headers. These reportedly flowed better than the standard manifolds even though interference issues with the starter resulted in a something of a compromised design on the passenger side. There were several versions of these manifolds depending upon the year and particular chassis the engine was destined for. In truth, most enthusiasts including myself ditched them for tubular headers which looked, sounded, and probably performed better. I've included some online photos comparing the standard and HiPo small block manifolds.

My versions of the HiPo manifolds will be CNC machined from single blocks of aluminum. During their design I also ran into starter interference problems, but unlike the full-size manifolds I decided to make both sides identical for symmetry's sake on a display engine. The internal passages will be created by pre-drilling the starting workpieces and plugging their open ends before machining. The manifolds were designed so they can be machined in just two setups using 3/8" cutters (and a little bit with 3/16" cutters).

The manifolds' modeling was yet another exercise in complex filleting, but thanks to the experience gained during the intake manifold modeling it wasn't as frustrating. As with the rest of this engine the goal wasn't to wind up with museum quality replicas, but they'll have the look and feel of HiPo's to anyone but a show-car nerd. - Terry

After some sketchy band sawing, the 2"x4"x10" starting workpieces for the manifolds wound up being the last of the large chunks of my 25 year old 7075 treasure find. There's no particular advantage in using 7075 for the exhaust except that it was the only material I had on hand that met the manifold size requirements. I checked Speedy Metals' pricing for back-up stock just in case I manage to screw these up, but their price out the door for the same two pieces of 7075 was close to $700. In comparison, 6061 is one-third the cost. I've been using 7075 exclusively for this build with little idea of its value and now really feel bad about the huge piece wasted when my carelessness scrapped the first block.

The drilled holes for the manifolds' internal passages required references to the faces of the workpieces, and those workpieces will be flipped and rotated in later setups. In order for the finished surfaces to smoothly blend together, it's important for the workpieces to be precisely square. To minimize errors, and since the outside dimensions of both manifolds are identical, the workpieces themselves were machined to identical dimensions. This turned out to be more important than I'd imagined when cross-checks between the two caught two errors that would have resulted in me sending money to Speedy.

The ports were pilot drilled and finished using a long ball end mill. The outer ends of the holes connecting the ports to each collector received plugs and were reamed to provide fine surface finishes for the Loctite. Fresh Loctite 620 and an activator were used to fill the .002" gaps between the parts. The plugs themselves were doubly secured with Loctite'd aluminum pins. After the Loctite cured, the collectors were opened up by step drilling. My mill didn't have enough head room for a drill chuck, and so the drill sizes were selected based upon what I had for R8 collets. The collector i.d. began at .312" (the port diameters) and increased to .750" at the collector flange.

After some 30 hours of preparation, the six pound workpieces are finally ready to spend time on the Tormach which will remove everything from them that doesn't look like an exhaust manifold.

I'd have gotten further along but we were hit with one heck of a hailstorm last week. In the last photo my wife is holding an example of the stones that pelted us for some ten minutes. The house and shop roofs and gutters will now have to be replaced, and both of our vehicles are likely total losses. Half the week was spent hauling away nearly 1000 gallons of yard debris. There's nearly $50k in casualty loss, but fortunately most of it is covered by insurance. - Terry

From a machining perspective the manifolds are organic parts with no convenient work-holding surfaces and so were designed to be machined through the front and rear faces of a rectangular workpiece. The workpieces included a framework around each manifold to support the semi-finished parts during their machining. The majority of each manifold's machining was completed through the front face of its workpiece while the manifold was still minimally attached to its rear face. Quick-set epoxy poured around the semi-finished part firmly reattached it to the front of its framework. The workpiece was then flipped over, and the part finish machined through its rear face while supported by the epoxy. Devcon 5 Minute epoxy which has a 200F service temperature was used, and after all machining was completed the finished manifolds were freed from their frameworks after a three hour 275F oven bake.

Most the chips removed from the workpieces were taken through their front faces in three roughing operations using a 3/8" end mill. These operations left .005" excess material around each manifold for finishing with a 3/8" ball mill. Two more minor operations with smaller end mills wrapped up the six hours per manifold frontside machining.

Before pouring the epoxy the earlier drilled mounting holes were blocked with silicone plugs commonly used in powder coating. An overnight cure minimized any gumminess that might clog the flutes of the cutters. The huge workpieces got incredibly hot (145F) while curing.

Using the same workpiece corner to reference all front and rear face operations ensured a best possible registration of the finished surfaces. Four flute cutters were used for the front face operations, but two flute cutters which better handle resins were used for the rear face operations. The necessarily slower feed rates doubled the machining times of the rear face operations, although total machining time for the rear side worked out to about the same six hours per manifold.

The ends of the workpieces were sawed off, and after an oven bake the finished manifolds were free of epoxy. Good results were obtained by peeling the epoxy from the manifolds while wearing gloves and the parts still hot. Stubborn patches of epoxy were scraped away hot using a wooden stick. I was greatly relieved to find all the plugs and their pins invisibly blended into the finished surfaces especially since the plugs wound up in filleted areas without a lot of overburden. Before claiming success, though, the manifolds were heat cycled several times.

The finished manifolds were trial-fitted to the block to verify starter clearance as well as clearance for a spark plug socket. This required fabricating a bell housing adapter for my little engine stand. The manifolds were then bead blasted and coated with Eastwood's high temperature manifold paint. The parts were allowed to dry for 24 hours before being baked for a couple hours at 300F to complete the paint's high temperature cure. The Eastwood paint is a much better match to a cast iron surface than was the brushed stainless Gun Kote used earlier on the bell housing. And so it was re-coated at the same time.

After a couple more cosmetic parts I hope to get on with the main bearings. - Terry

Valve cover t-bar hold-downs were popular dress-up items on engines from the 289's era, and a set for the model was machined next. The full size versions were chrome plated, but the model's were machined from polished stainless steel. Their machining was straightforward and included a handful of simple lathe operations. The two piece assemblies were pressed together with the help of a simple shop-made fixture.

A pair of oil filler caps was also machined for the rocker covers. The indirect paths connecting them to the sump would make filling the sump through the caps painfully slow, and so it will instead be filled through a special port on the pan. These paths however are more than adequate to allow the crankcase to breathe through the caps. The two-piece cap assemblies were designed to flow crankcase gasses and hopefully encourage an oil film to settle on the valve train components under the rocker covers.

In order to reduce machining time and material waste during the intake manifold's machining, the carb adapter wasn't machined as an integral part of the manifold. Instead, it was next machined as a bolt-on item. As mentioned earlier, the manifold's throat is surrounded by a well that may be used as a fuel reservoir for a return-type fuel loop planned for the engine. Carburetor details still need to be worked out, but the well was extended into the adapter and remains an option for the reservoir should carb bowl(s) not prove practical. In any event, a .015" thick Teflon gasket seals the adapter to the manifold.

A batch of hose clamps wrapped up the last of my excuses to start work on the main bearings. The already machined barbed fittings for the coolant system were designed for 3/8" i.d. (1/2" o.d.) black silicone hose, and so a half dozen or so billet clamps were machined around that particular hose and those barbs.

The clamps are secured with integral 2-56 SHCS's. The optimum body i.d. using a .063" wide slit was experimentally determined to be .010" greater than the hose i.d. The clamp bodies were machined in a single cookie sheet batch, but finishing required a couple secondary drilling operations in addition to one with a slitting saw. Their surfaces were finished with a one minute dip in an aluminum brite product. - Terry

Hi Terry
Outstanding as always! A real pleasure to see your work, can't thank you enough for your detailed posts and pictures.

If you need another excuse to put off the main bearings you could always machine the bottoms of the T-Handle clamps so all the handles "clock" in same direction :) or use a rubbery washer under them to get a little leeway on where they stop.

Just kidding really, truly marvelous work !!!


Anxiously awaiting the next instalment :)

Your CNC machined parts look like sand cast after the bead blasting.

I've probably asked this before, but... What's your typical stepover and depth of cut on the final pass? In particular, the corners on the exhaust header where it bends after coming out of the head, did you do any sanding there? (You said you had some scraping and filing to do.) Or were they just machined, then bead blasted?

I do believe that you've perfected the process for simulating sand cast parts in miniature.


Your CNC machined parts look like sand cast after the bead blasting.

I've probably asked this before, but... What's your typical stepover and depth of cut on the final pass? In particular, the corners on the exhaust header where it bends after coming out of the head, did you do any sanding there? (You said you had some scraping and filing to do.) Or were they just machined, then bead blasted?

I do believe that you've perfected the process for simulating sand cast parts in miniature.

By scraping, I meant there was some epoxy left in some of the corners that I needed to scrape out with a sharp stick. Bead blasting doesn't do a good job of removing epoxy as it seems to bounce off it without cutting. The finishing passes were set for a .0005" maximum scallop height, and those passes were working with .005" material left from the previous roughing passes. There were two small areas near the collector flange that needed work with a needle file that my cutters couldn't reach. Other than that, the manifolds went from the Tormach to the bead baster with no other filing or sanding. - Terry
Last edited:
Thanks Terry.

I'm thinking about casting some parts for some RC truck and construction equipment projects that I've got lined up as future projects. These would be cast using either lost wax or lost PLA 3D printed patterns. I was wondering if I could use your bead blasting process to eliminate the layer lines from the cast parts without any other post-processing. It sounds like that will be do-able as long as I print using 0.1mm layers. (That's about 4 thou-ish for the Metrically challenged.) My printers will happily print at that layer height.

I think you said that you are using glass beads from Harbor Fright, is that right? (What grit/size are the beads?)

Thanks Terry.

I'm thinking about casting some parts for some RC truck and construction equipment projects that I've got lined up as future projects. These would be cast using either lost wax or lost PLA 3D printed patterns. I was wondering if I could use your bead blasting process to eliminate the layer lines from the cast parts without any other post-processing. It sounds like that will be do-able as long as I print using 0.1mm layers. (That's about 4 thou-ish for the Metrically challenged.) My printers will happily print at that layer height.

I think you said that you are using glass beads from Harbor Fright, is that right? (What grit/size are the beads?)

With my unknown grit Harbor Freight beads, I can usually bury .001" machining marks, and a little more if I'm willing to dwell long enough. I wouldn't be able to hide .004" though. What happens is that on a finished surface with .0005" scallops directly off the mill, the surface is so reflective that defects are difficult to see, and if a polished surface is wanted, then you're finished. With the matte finish of the beads, though, every defect clearly shows up. That's why I try to get the best finish possible on the mill even if I'm after a cast looking surface when I'm done because once I start filing, a bead blasted surface is unforgiving and will need time to make it right with a file. - Terry
Last edited:
Miss-read the 0.0005" height as 0.005" - my bad. (I knew it sounded too good to be true.)

I may have to invest in some Poly-cast filament and do some IPA vapor smoothing on the printed pattern. The printer specs SAY it'll go to 0.05mm layers but I'm taking that with a grain of salt.
The crankshaft mains include a pair of outer ball bearings and three inner bronze clamshell bearings. The diameters of all the bearings are the same, and the recesses for them were machined into the block using the Tormach. For all practical purposes they're equivalent to having been line bored. With just the ball bearings installed, a test shaft spins freely in the block with essentially no runout. With the crank installed for the same test, the runout measured at two journals is a thousandth and approaches .002" at its nose.

The bronze bearings were next on the list of parts to be machined. A crescent cut through the side of each bearing will allow oil to enter the bearing and be gravity fed into a groove that will supplement the bottom end's splash lubrication.

A length of 2" diameter 932 bearing bronze left over from an earlier project was selected for a workpiece. Although it was just long enough for three bearings and a spare, its diameter was considerably greater than needed for 1-3/8" bearings, and so much of it was destined for the chip tray.

The first step was to saw the material into two lengthwise halves. A carrier fabricated from a piece of square steel tubing was used to support the material while it was being band-sawed. The sawed surfaces were cleaned up on the mill and the two halves rejoined with super glue. Using a spindle microscope, both ends of the workpiece were center drilled on the mill at the midpoint of the parting line. The workpiece was then set up between centers on the lathe where it was turned down to its finished diameter of 1.3745".

The workpiece was returned to the mill and clamped in the vise with the parting line horizontal. The holes and counterbores for all four bearings' mounting bolts were then drilled and temporarily tapped. Burrs raised by the drill when it exited the curved rear surface of the workpiece were removed by slightly counterboring the opposite side.

This was the first time I'd used super glue in a machining application and didn't yet trust it. And so, with all eight threaded fasteners installed, the workpiece was chucked in the lathe's 4-jaw and bored .003" over the diameter of the crank journals. Half of this clearance was allotted for an oil film, and the other half was to account for the crank runout and accumulated machining errors. Individual bearings will be scraped as needed later once the crank is installed. A sanity check with the test shaft spinning freely in the ball bearings and the semi-finished workpiece temporarily resting (or with downward pressure) in the block seemed encouraging. - Terry

I would have thought that the pressure from the centers would have wanted to split the halves apart along the glue line, like a wedge. Since it obviously worked what was your secret? Minimal pressure on the centers and lots of glue area? What was your backup plan if it had split?