Ford 300 Inline Six

[mayhugh1]

A bell housing cover plate was designed and machined to fully enclose the flywheel. After broaching the flywheel for the crankshaft key, the 83 tooth ring gear that I'd hoped to become part of a functional starter was covered up. I staged a photo of one of my 390 test motors sitting in the bell housing as a memento of the weeks spent trying to bend reality. Work then began on it cosmetic replacement that turned out to be rather tricky to machine.

The motor's 3/4" diameter main body was turned on the lathe, and the details on its end were machined on the mill. Its distinctive bendix cover contained a lot of filleting and was machined as a separate part of the starter. Its complex shape left little usable area for work-holding while its bottom concave surface was machined. It was then permanently joined to the body with a single Loctite'd screw.

As shown in the photos, the motor received three different surface finishes: bead blasting, bare polished metal, and black Gun Kote. Remarkably, and quite by accident, the exhaust pipe assembly made earlier fit nicely around the starter. - Terry

 

[ddmckee54]

When you said "it's held on with a single loctited screw", I assume you were talking about the Bendixx cover. How is the faux starter motor held in place, since it has no visible means of support?

 

[mayhugh1]

When you said "it's held on with a single loctited screw", I assume you were talking about the Bendixx cover. How is the faux starter motor held in place, since it has no visible means of support?

There's a hidden 8-32 flat head screw through the cover plate and into the rear of the motor's main body. - Terry

 

[ddmckee54]

I thought it was probably something like that. Scale studs and nuts, with a scale flange on the starter would have been slightly over the top.

 

[mayhugh1]

I thought it was probably something like that. Scale studs and nuts, with a scale flange on the starter would have been slightly over the top.

Actually, I would have added the mounting flange, but I'd designed the cover plate to handle both the 390 and faux motors. The brush motors all attach with a pair of screws in the end of their front face and didn't require a flange. I guess I thought I might come back with a 390 solution someday. - Terry

 

[ddmckee54]

You've probably already covered this, but did you look into using a motor with a planetary gearbox attached for the starter?

Other than the fact that it would be about twice a long, look REALLY out of place, require an actual Bendix or some type of over-running clutch.... Yeah, nevermind.

 

[petertha]

Not even sure it would work, but I had a look around some RC gear drive motors I'm familiar with. This one is one of the smaller ones, 320 watts on 6.7:1 reduction, The motor can is 22mm (0.87") but the gear drive about doubles the overall length. They are spendy. Also brushless, so you are into a specific (RC) speed control unit although the voltage might be a good match (3S lipo = nominal 12V).
https://www.hyperflight.co.uk/products.asp?code=PLM-1010-19&name=powerline-micro-1010-19-68g
I've seen some gear reductions on the brushed motors similar to what Terry has, but the power was significantly less. For reference I think cordless drills might be in the 500W range? depending on cell count etc. Hard to miniaturize that by a large factor.

 

[Peter Twissell]

Brushless motors are not well suited as starters.
The controller will try to ramp up the motor speed and if the motor doesn't keep up, it gets out of synchronisation with the electronic 'commutation' and produces no torque.
This can be overcome if the motor is fitted with an angle encoder, so the controller can match the wave form and frequency, but everything then gets more expensive again.

 

[petertha]

That's interesting Peter, never considered that. Not to detract from the build thread but how is that different to the motor throttling the propeller load up & down & holding anywhere in between like it normally does? Do you mean once the engine has started it will have potential to over-drive the brushless commutation timing, or just these kinds of motors in general?

 

[Peter Twissell]

Brushless controllers which run "open loop", without motor position sensors, are programmed to limit the rate at which motor rpm changes.
Ann inertial load, such as a propeller, will allow the motor to pick up speed in direct proportion to the torque.
A starter is loaded in a less predictable way and may well be effectively stalled when power is first applied.
If a sensorless brushless starter were used, when the engine starts and tries to overdrive the motor, the motor will act as a brake until the Bendix disengages.

 

[mayhugh1]

Before starting on the bottom end assembly, the motor mounts were machined to support and help steady the block. A baseplate will come later after I've finally decided between a fuel pump and a conventional gravity fed tank. The mounts could have quickly been made from angle stock, but with none on hand they were wastefully machined from thick bar stock. After bead-blasting they were gold anodized using the home anodizing setup described in my Offy build. While the chemicals were still laid out, the bell housing was clear anodized to darken its color and to reduce the chances of oil and grease stains later on.

But, the anodizing didn't go smoothly this time. The two front mounts turned out great on my first try, but my beginners luck ran out on the rear mount. Initially its color was off just a bit, and not knowing any better I decided to re-do it. After again bead blasting and etching it in lye, I wasn't able to get an anodizing current to flow through it. I expected the problem was likely a poor electrical connection to the part which in my experience in an ongoing problem with this process. However, nothing I did including replacing the electrolyte and wire brushing the lead electrodes would reestablish current. The unregulated power supply I'd been using had a maximum output voltage of some 24 volts, and my typical 1-2 amp anodizing current was well within its capability.

The part's surface was redone a third time and with multiple electrical connections but still no joy. Finally, switching to a constant current supply with 2X more output voltage solved the problem, but strangely it had to be cranked up to nearly 50 volts to get 800 ma. of anodizing current. After 20 minutes in the tank I noticed the part had taken on a slight golden tinge similar to its original color. After the color dip and sealing steps, ohmmeter measurements showed the part was finally anodized, and its color was as nice and uniform as I could ask for.

I couldn't find any online discussion about what I'd run into, but I had a suspicion the subsequent bead blasting's hadn't removed the entire initial anodized layer after all. Although my ohmmeter measurements had indicated zero surface resistance, my probes were making contact to only a relatively few microscopically high points that had actually been cleaned off by bead blasting. This tiny effective area wasn't indicative of the entire part's surface, and the very thin insulating islands remaining between them required a high punch-through voltage in order to get significant current flowing. After thinking about it, I was amazed that the part turned out as well as it did.

During assembly, the ring'd pistons were attached to the crankshaft through the top of the block. The shop-made installation tool shown in one of the photos was used to eliminate potential damage to the rings. The dimensions of its two-piece design aren't critical, but the i.d. should closely match the piston's diameter. An o-ring held the halves in place around the piston and the rings squeezed down inside their grooves so the piston could be safely slid into its cylinder. The only tricky part was lining up the rods with their journals inside the narrow spaces between the crankshaft webs.

With the crank fully assembled, the block was sealed to the pan with a two-piece .015" thick Teflon gasket set. Adhesive backed vinyl was used to seal the pan to the front and rear main (ball) bearings. The shape of the flywheel along with the rings' friction will make it awkward to precisely position the crankshaft during timing adjustments. So, the adapter that will eventually be used to drill start the engine was machined next since it can also be used to manually rotate the crank.

The drill starter consists of a machined body into which a one-way clutch was pressed. The inner bearing for the clutch was machined from drill rod and then hardened. Its hex shaped end matches the socketed rear of the flywheel. The inner bearing is loosely retained to the body with a 6-32 SHCS. - Terry

 

[a41capt]

Amazing work Terry! I’ve given home anodizing a try in the past and I’m setting up to anodize the parts on my Webster. First time I’ve tried to pretty up an engine!

John W

 

[ddmckee54]

I liked your ring compressor. The bell housing and engine mounts look good too.

 

[a41capt]

Terry, what are you using as a sealant for the dye on anodized parts? Do you just use the boiling water method, or are you using a commercial sealant?

Thanks in advance,
John W

 

[mayhugh1]

Terry, what are you using as a sealant for the dye on anodized parts? Do you just use the boiling water method, or are you using a commercial sealant?

Thanks in advance,
John W

Just 20 minutes in boiling water...

 

[a41capt]

Just 20 minutes in boiling water...

 

[Bentwings]

Nice work . I’ve wanted to try this for a long time I’m making an aluminum electric boiler for my steam engine I’d like to use som an fittings they are blue red or black so I could make the boils som cool color too . I have not done much anodizingbin industry . The boiler is kinda high tech being aluminum I’m under fire for deviating from traditional copper in modeling but copper has structural limits that need to evconsidered . Aluminum can handl pretty high temps as long as you stay in limits my unit is more direct fired so I don’t have that issue copper works ok too but cxcess heat and raw fire ca be dangerous so they have closer guide lines I’m more or less on my own as few build electric in the first place so strict rule I have to us material properties anodizing defects don’t seem to be well published . I’ve not had any issues with anodized parts in race cars boats still get coorosion but with proper anodes it goes away I’ve had a lot of hard anodizing done in industry snd racing and it worked well for where it was used
Byron

 

[Peter Twissell]

Byron, I'd be very cautious with an aluminium boiler.
Aluminium loses a lot of its tensile strength as temperature rises and is prone to fatigue.
Anodising makes things worse, as it introduces a brittle surface which in turn generates crack initiation points.
Copper is relatively ductile, so in the case of a boiler failure, is more likely to open a leak than split apart catastrophically.
Aluminium has about half the stiffness of copper, so it will move more under pressure, putting additional stress on joints.
I have used aluminium bodied hydraulic accumulators, but they are wrapped in carbon fibre to overcome the stiffness issue and are rated to 120C maximum temperature.

 

[mayhugh1]

Before installing the head, the crankshaft was rotated to TDC of cylinder number one and its position memorialized with a mark scribed on the crankshaft pulley next to the timing indicator. This mark will be used later to adjust the distributor timing just as it's done on the full-size engine.

The head, along with the Teflon head gasket made earlier, was then installed using fourteen 5-40 steel SHCS's. Standard length screws were shortened to obtain maximum possible purchase (6 threads) inside the block.

At this point it was possible to leak check the entire coolant system using a vacuum test. Unfortunately, it failed miserably. An easily fixed leak was discovered at the water pump where an interference with the timing gear housing prevented it from fully seating on its gasket. However, a more serious problem remained. Eventually, a leaky liner was discovered after removing the pan and flooding the bottom of the block with oil. With the coolant system now pressurized, bubbles could be seen escaping from the bottom of the liner in cylinder number two.

Although some builders use press-fit liners that are honed after installation, I prefer to completely finish mine outside the block. Since they're machined with close sliding fits inside the cylinders, they must be sealed with Loctite. With the 3/16" glue surface left around the bottom of each liner, I was surprised to discover a leak.

Resealing the liner began with a solvent flush of the coolant system to remove any traces of oil that might have been drawn into the leak during the vacuum test. After removing the head, the block was set on a pair of wood blocks with the bottom facing down. Any remaining solvent was allowed to drain through the leak for the next few hours. Using a needle syringe and working through the two transfer passages on either side of number 2 cylinder, Loctite 290 (a wicking grade) was squirted into the space surrounding the liner. The Loctite began leaking through the liner half an hour or so later. Clean paper towels under the block helped keep track of the sealing progress.

Several hours later, Loctite 609 (slightly more viscous press fit) was added. The draining slowed greatly after a few more hours, and then the block was set inside my home-made welding rod oven for an accelerated cure. With an inside temperature of 140F, the block was allowed to cure overnight.
A couple pressurized solvent flushes were used the next day to remove any uncured Loctite remaining inside the block. After reinstalling the head, the vacuum test was repeated showing the coolant system was finally leak free.

The next step is to machine and install the lifters and pushrods, but I only just discovered I'll have to order the material for them. - Terry

 

[ddmckee54]

30cc's coolant capacity, wonder what that is in 1/5 scale gallons?