1/3 Scale Ford 289 Hi-Po

The camshaft was tackled next. Ninety degree heads, offset cylinders, interleaved pushrods, firing order, direction of rotation - all these combine to make the scratch design of a V8 camshaft a challenging mental exercise.

First up was the selection of the camshaft's duration and lobe separation angle (LSA) which pretty much define the engine's starting and running characteristics. I decided upon a mild street cam with 260 degrees of duration and an LSA of 112 degrees and then created a model of its intake/exhaust lobe pair. Since the left and right banks require oppositely ordered lobes, two different pairs were created. While being careful to assign the correct lobe pair to each engine bank, the firing order was used to sequentially arrange the lobe pairs with each intake lobe being retarded 45 degrees from the one before it. A machined flat on the nose of the camshaft defined the arbitrary starting angle.

The flat tappet cam originally designed for this engine was modified to take advantage of the roller lifters. Since this was my first roller cam I was a bit conservative, but when compared with the original cam it will fully open the valves a little sooner and keep them fully open a little longer for the same duration and LSA.

The starting material for the camshaft was a nine inch piece of 5/8" ground/polished Stressproof. Its .624" measured diameter fit perfectly inside the block's .625" inner bearing bores, and so its bearing surfaces didn't require machining. The lobe blanks were manually turned on the lathe before the workpiece was moved to the Tormach where they were machined into lobes using a 4-axis rotary operation.

The camshaft was divided into four machining zones with two lobe pairs in each zone. The blanks were turned and the lobes machined in one zone before moving on to the next. A heavy steel disk set-screwed to the flat on the nose of the camshaft was used to re-reference the workpiece in each of the four milling setups. It also damped some of the machining vibrations. A shop-made fixture stabilized the portion of the camshaft sticking out of the fourth axis.

Each lobe was machined in three operations using a 4 flute 1/8" end mill with a .010" corner radius. The first two operations were roughing operations at fixed A axis angles of 0 degrees and 180 degrees. These operations left .005" excess stock for the rotary finishing operation. Total machining time per section was just under two hours.

The relief that's ground into the bottoms of cylindrical end mills left 'furrows' in the finished lobe surfaces. These marks don't happen when an end mill is normal to the surface it's cutting such as when a flat surface is machined. However, in order to prevent gouging in the rotary operation, the cutter was held slightly tilted to the lobe's continuously moving curved surface. Fortunately, the furrows were easily polished out with a strip of 800g paper glued to a flat stick. The marks were polished out while the camshaft was spinning in the lathe at about one rpm to reduce chances of inadvertently modifying the profiles.

Upon finishing the lobes, the rear of the camshaft was machined to accept the rear ball bearing and a block cover plate created for it. This completed the camshaft except for the custom helical gear needed to drive the distributor. That gear will become its own project later. For now, the valve cages seem to be the most logical next step. - Terry

Fantastic work !

This thread is golden.

.

Wow so beautyfull

How are you driving the distributor? I don't see a drive gear.

Hi George
He stated in the last couple of sentences
"This completed the camshaft except for the custom helical gear needed to drive the distributor. That gear will become its own project later.

Scott

Sorry I didn't read far enough

My last photo showing the 'assembled' camshaft may be misleading. The camshaft parts was really just stuck together for the photo. The body with all the lobes is separate from the front hub. These two pieces have tenon'd ends which will be Loctite'd together inside a sleeve which will be machined into the helical drive gear. The camshaft drive sprocket is also a separate part and is bolted to the front hub through slotted holes which will allow it to later adjust the timing. - Terry

It's difficult to separate the design and machining of the valves from those of the valve cages. In addition to sealing the combustion chambers they affect their volumes (and C.R.) as well as the rocker arm geometries and pushrod lengths. Even with their CAD models playing nicely together, subtle differences in their machining can even prevent a model engine from starting. Iterations of both the valves and cages were made and tested together.

A V-8 needs lots of valves and cages, and so another goal while developing the trial parts was to come up with the machining steps that will be needed later to produce them consistently in large numbers.

The valve cages are being machined from 544 phosphor bronze which has been my goto cage material. Their design is simple enough, but getting them to seal can sometimes be tricky. The most important operation is drilling and reaming the hole for the valve guide. Unlike the throat of the cage which is being bored, the drilled hole for the valve guide isn't guaranteed to wind up precisely on the axis of the cage which is important for the seal. A manual seat cutter which is being used cut the actual seats is piloted by the valve guide.

The finished cages will be leak-checked with a test valve before and after installation. To prevent distortion during installation, the cages are being machined for slip fits in the heads and then installed with Loctite 620.

The side cut-outs for the ports are usually drilled after the cages are installed. Unfortunately, the trajectories of the ports in these heads brought the cut-outs for the intakes uncomfortably close to the seats. Because of this the cut-outs are being drilled (using a 2 flute v-cutter) before the cages are installed so any distortion created by their drilling will be caught by the leak check. Even though there were no issues with the exhaust cage cutouts, the exhaust cages are being treated similarly.

The valves are being machined from 303 stainless which is the material I've used for the 100+ valves in all my other engines. Except for the two-part brazed method, I've tried most of the popular valve making techniques that have to deal with the unwieldy stems. What works best for me is to divide the length of the valve's workpiece into several short consecutive machining zones, and complete the machining in each one before proceeding to the next.

Since a test valve was needed, it became a 'deliverable' of the valve development. For consistency, the blanks for the valves are being machined on my little Wabeco CNC lathe using the multi-zone technique described above. After several trials I had a couple first articles that I was happy with and felt could be consistently reproduced. One of the trial valves was converted to a test valve by drilling a .040" hole lengthwise through its stem to allow a vacuum to be pulled through it.

Several cages were made and leak-checked using the test valve. The valve cage blanks were also turned on my Wabeco. The seats were cut using a 45 degree piloted muzzle reamer purchased from Brownells several years ago. Over time the technique I've developed for using this seat cutter with phosphor bronze is to hold the cage vertical and allow only the weight of the cutter to apply the cutting force. The flutes of the cutter are coated with motor oil for the best possible surface finish. The seat widths generally wind up on the order of five thousandths which is all that's really needed. Going deeper with additional force risks marking the seat, and once marked the cage is usually scrap.

For the leak-down test a Mini-Vac is attached to the rear of the valve/cage assembly. My criteria which may be over-kill for a 'pass' is a 25 psi to 10 psi leak-down of 10-15 seconds. The last two trial cages met the low end of this spec. I don't lap valves with their cages. In my experience, lapping model engine valves usually transfers the problem from one part to the other, and the leak usually becomes worse. Manually polishing the cage's seat with an oil solution of extra fine Timesaver on a felt bob can often extend the leak-down times up to a minute.

Along the way a few shop-made tools were also made to help with production later on. After parting the valves from their blanks, two sets of split collets support the valves while their faces are being finished and their keeper being grooves cut. A Delrin installation tool was also made to align the cages with their ports while the Loctite sets up. The next step is to repeat all these machining steps another twenty or so times. - Terry

Thank you for posting all those pics of how you made the valves and cages ! Pics can fool the eye, what size are the valve stems and heads?

To me, it looks like the test setup does not test leakage between the valve stems and the cages, am I correct? Are stem seals of some sort planned?

Extremely beautiful work as you always do !

Sparky,
The valve heads are 3/8" in diameter and the stems are 5/32" in diameter.
If I understand your question, the leak check isn't trying to measure a component of leakage between the cage and the valve stem because it doesn't affect the main seal of the combustion chamber. I typically leave just over a thousandth clearance between the valve stem and the valve cage. In a model engine, fuel and oil vapors tend to find their way in between the two to lubricate the stem, and so no valve stem seals. - Terry

Terry I may have even asked this before but coincidentally just putting together materials order for my next engine. I also used 303 for my valves but at some point considered 416. According to McMaster abbreviated specs, 416 strength is 40Ksi vs 303 at 25Ksi. The machineability (simple bar comparison) seems the same as 303. 416 is slightly harder. I even bought a stick of 416 & confirmed it machines well. But they list it as (slightly?) magnetic if that's a factor. Any comments from your perspective?

petertha said:

Terry I may have even asked this before but coincidentally just putting together materials order for my next engine. I also used 303 for my valves but at some point considered 416. According to McMaster abbreviated specs, 416 strength is 40Ksi vs 303 at 25Ksi. The machineability (simple bar comparison) seems the same as 303. 416 is slightly harder. I even bought a stick of 416 & confirmed it machines well. But they list it as (slightly?) magnetic if that's a factor. Any comments from your perspective?

Click to expand...

Peter,
I may have worked with 416 but my memory is foggy and most of the material I use is recycled scrap and machine shop drops with no pedigree behind them. It's possible I have some 416 in my shop as I have some drops that are bright and shiny like 303 with no corrosion but are magnetic. The specs you posted look good enough that I'll likely give it a try - Terry

I've turned IC engine valves from 17-4 PH (precipitation hardening) stainless steel in "Condition A." I think Condition A is the way suppliers like McMaster Carr offer it. After turning and polishing, I heat it to 900°F in air and then air cool, yielding hardness around 40 HRC. The material is not difficult to machine in Condition A. It's certainly nowhere near as challenging as austenitic stainless steels. I don't think the valves will ever see 900°F once installed in a model engine.

Some twenty valves and valve cages were completed, and so there were plenty of spares. Cage leak-down times measured using the test valve were surprisingly consistent at some 15 seconds plus or minus a few. Ten seconds of polishing with extra-fine Timesaver smeared on a felt bob improved the times for a few cages to beyond a minute, but for the majority it added another 10 seconds or so. For a couple cages it made no difference at all. Drilling the ports with a v-cutter didn't seem to affect the times, although only spot checks were made due to the difficulty in wrestling a sealing sleeve over the cutouts.

The cages were installed in the heads with Loctite 620. The set-up times were scary short. There was barely enough time for the installation tool to align the port cutouts in the heads. In fact, the first cage immediately set up misaligned and had to be driven out with a drift. After an overnight cure, the installed cages were leak-checked using the newly finished valves. A silicone cap sealed each valve stem to the rear of the cage while vacuum was pulled through the ports behind the valve.

Final leak checks were made after installing the valves with their springs and retainers. A drop of oil on the valve stems effectively sealed them to the cages while vacuum was again pulled through the ports in the heads. Final results were acceptable but not as good as that measured with the test valve.

The rockers and roller lifters were finally assembled from the piles of individual parts machined earlier. After verifying their length, twenty pushrods were machined from unhardened drill rod. The Wabeco lathe was used to CNC machine ball ends on each rod.

The compression ratio estimated during early the engine's early modeling used SolidWorks' .174 cubic inch measurement for the volume of the combustion chamber. This measurement however didn't take into account the space occupied by the valves and spark plug. These effects are typically negligible, but since the c.r. calculation also showed a full point sensitivity to the .020" head gasket, the combustion chamber deserved a closer look.

With a pair of valves and a spark plug installed, one of the combustion chambers was cc'd so its true volume could be used in the c.r. calculations. For those unfamiliar with the term "cc'ing", the combustion chamber is sealed with a plexiglass cover plate that has a hole drilled through its center. The chamber is then filled through this hole with kerosene from a burette which is essentially a calibrated dropper. This measurement showed the actual volume is 2.65 ml or .162 cubic inch which was a 7% reduction in the combustion chamber's volume. In my case this will add an extra half point to the compression ratio which is significant but less than expected.

It was no surprise to find near zero clearance between a fully opened valve and a piston at TDC since the pistons in the full-size engine were eye-browed. In the early modeling the wedge style heads were creating interference issues with the placeholder flat-topped pistons. The next steps will be to finish machine the pistons for appropriate valve clearance and a final compression ratio of 6.5 to 7.0. - Terry

Hi Terry,
Your build is a lot more technical than mine was. I never CC'd my chambers. My closest guess would be that the C.R. came out to about 8:1, without the head gasket so maybe 7-7.25 with the gasket. Have you made head gaskets yet? If not let me know and I will send you the information on the material I used. It has held up for over 10 years with no leaks whatsoever.

As an aside. I knew the small block Fords had separate rocker arms but with my engine having a pressure oil system I decided to go with a rocker shaft like the Ford Y blocks and FE engines. I ran a small copper tube from the main oil galley that runs parallel to the camshaft up through the head and into a fitting on the end of the rocker shaft. I don't know how much oil it was feeding but every time I took the rocker cover off there was plenty of oil in the head. After awhile I decided to take the feed line out to make sure the vital parts had enough oil pressure. When I pulled the rocker covers off there was always a coating of oil on everything from the oil vapors that would come up from the engine and vent out of the breather caps. After a show I always pull the covers and put a drop of oil on the rocker arms and have never had any issues.

The compression ratio using the measured combustion chamber volume worked out to be 6.9. I was hoping for a little more so the pistons could be simply dished to provide the needed valve clearance. At five to ten thousandths interference the goal was a clearance of .050" to insure there would be no interference even at a lift that would fully compress the valve springs. Instead, a pair of eyebrows were cut into the piston tops since they will add minimum volume back into the combustion chambers. Because of the engine's symmetry, both banks can use the same pistons so long as they're installed in opposite facing directions. There was just a negligible effect on c.r. which dropped to 6.8.

The fixture used earlier to hold the pistons while their wrist pin holes were being drilled was resurrected for the eyebrow machining. They were cut by simply plunging an end mill that was .045" larger in diameter than the valve heads. Before modifying the actual pistons the setup and machining were verified on a scrapped piston which was used to sample the clearances in both banks.

The next step will be to finish the camshaft which still needs a custom DP helical gear set to drive the distributor. My last experience with cutting helical gears came with a long and frustrating learning curve, and so I hope I'll be able to resurrect the process from the notes and posts made during my Inline six build. - Terry

I made a set of helical gears for my gatling gun build using Gearotic Motion software on my cnc mill. First time using gearotic and first try the gears came out great. Made with just a 1/16 straight endmill. I believe there is a free trial version.

I am a hack compared to your beautiful work. I much enjoy following and admiring your work.

mayhugh1 said:

At five to ten thousandths interference the goal was a clearance of .050" to insure there would be no interference even at a lift that would fully compress the valve springs.

Click to expand...

That's a good way of thinking about allowing for the maximum displacement of the valve - it cant go any further than that into the head. When I cut the eyebrows (learned a new word) into my pistons, it was addressing small but unmistakable valve marks on the piston after the first run. Lucky me it didn't do catastrophic damage. All my machining dimensions checked out according to plans so I chocked it up to the grey area of running conditions: parts thermally growing, or maybe throwing the valve under inertia slightly beyond what the cam was theoretically driving it. I didn't have a lot of piston crown thickness to work with so it was an educated guess of what would work & not risk a crack in the piston on the sharp relief line. Next step would be new angle topped pistons or new cam plates or similar.

For future build consideration, have you ever heard of an allowance rule of thumb like make allowance X% of cam lift or Y% of valve length or...?

the secret to success with helical gears is to design your engine around commercially available ones !!! (hpcgears.com has the smallest I know of at 15-T 48-DP 45-deg, which I'm using in both my Duesenberg and Hansen engines)