Forrest Edwards radial 5

Hi there!
Following my thread in the welcome section, I started to explore the possibilities for the build of the Forrest Edwards radial, in a slightly downsized version. As explained in the aforementioned thread, I intend to use some bike hub gears for the cam drive. In order to see if this works, I made a mock-up withe some scrap metal to chweck the center distances of the gears.Now there is a problem when downsizing this engine, I intend to downsize to 80% of the original size. The shaft of the idler gear in the original design is supported at one side only, in the crankcase. I want to fit the idler gears on a shaft, supported on both sides withe miniature ball bearings.I put it al together and it works rather nice, no excess backlash and no binding.[/ATTACH]

So, up to the next step.I will make a wooden model to get an idea of the size of the machine and to try some design changes.

To be continued....
Jos

 

In the winter I have been working on a wooden model of the downsized (80%) Edwards.I do not have milling equipment so I try to make as most parts as possible by turning. This is not a big problem because most parts that need a milling operation ( conrods, cylinder head cooling fins,rocker arms) can be made on the toolpost of my lathe (Emco Compact 8- Austrian made). I started withe the crankcase,made of two layers of 1 mm plywood wrapped around a piece of steel tubing , glued with polyurethane glue and clamped withe two big hose clamps.Pieces of 3 mm plywood were glued on the circumference and a 8 mm plywood ring was glued inside. This engine will have a two-part crankcase, partly to reduce the risk of machining erors, partly because it is easier to machine and ,most of all, the state of raw material. Having worked 17 years as a QA engineer in the manufactring industry, I collected a vast amount of rejected parts, left over blanks, obsolete bar stock etc. I have excellent material for the crankcase bus it has a hole in the middle...Apart from this, machining the bearing seatings for the idler shaft at the exact center distance can be a tricky operation so it is easier to do this with a separate bearing cover than with an almost completed one-piece crankcase. So, after glueing an turning the inside of the crankcase I fitted it to a fixture made of a 90 degree block fitted on a flat plate clamped in the chuck and additionally secured with a long stud through the main spindle.This has the advantage that planing and boring of the cylinder hole can be done in one operation. I made a drill jig for the cylinder boltholes which can be used for the crankcase, cylinder barrel and cylinder head.If small inaccuracies occur ( which are very likely) they will be all the same in the mating parts so assembly will present no problems.
I used camgear parts from the mockup to fit on the front end.
3 mm bolts for the cylinder assembly turned out to be a bit lumpy so I will use 2,5 mm ones for the actual engine.
So foar, so good.. In the meantime, I have been working on a rotating table for the camring which is a major headache for me but I will show wha t ||I have made in a separate post.
For now, it' s holiday time so I will continue the work on the 0,8 Edwards in August or so .

Jos


Failure is always an option Mythbusters

 

Now it's been a year or so since I started preparations for the build of the downsized Edwards Radial 5 so , with the snow falling down in Holland, it's time to give an update of the things I've been doing the past year. I have been making jigs, fixtures, templates ,wooden test pieces and for all, a lot of re-calculating dimensions with the help of an Excel sheet for the downzizing and the conversion to metric.
By far the biggest headache was the fabrication of the camring which required an accurate index disc. I started withe the purchase of a rotating table to fit on the cross-slide of my lathe after removal of the toolpost. First I tried printing a tiing disc and gluing it to a 2 mm steel disc but the printing quality was far too inaccurate to mak this work. Browsing through a drawer with old stuff I found a meccano gear with 120 teeth. I made a construction on the rotating table with a spring-loaded holding pin to keep the gear in place. On top of the gear the disc was fitted. The holes were drilled in three different pitch circle diameters otherwise the holes would be too close together. So I found myself drilling 360 1mm holes. A tedious job but someone has to do it....

 

Engine built in wood, exciting to see if it was possible to run a motor made of wood.

Smart to make parts in wood and try to sample before everything is made in metals.

 

So, after the drilling was complete, I had to find a way to mark the index holes in such a way that the chances of making mistakes are minimal, bearing inmind that a small mistake could ruin hours of work to make a blank for the camring. I had a lump of nice steel, 42CrMo4, 80 mm dia., used in marine propulsion systems. But with no fancy bandsaw available, I had to cut some grooves with the parting tool after reducing the diameter. Then I drilled a row of holes in this grooves and cut off the remaining material with a handsaw.

I developed a system with colours marking the degrees on the index disc.
These colours correspond with colours on the lift chart.

To rotate the disc I used a worm gear ( Yes, Meccano again) which fitted on the gear previously used for indexing. A stop was fitted on the base of the rotating table . I used an 1 mm index pin on the disc , held in place by a small magnet.

There was another issue, being the need to move the cross-slide with increments of 0,01 mm. This required the use of e DRO, which I didn't have.
But I did have a digital caliper, so with a few fixtures it was bolted on to the cross- slide, and worked perfectly.

It seems common practise to use a woodruff key cutter fot the machining of the camring but these cutters are notoriously expensive. I used a standard end cutter ( cost about 7 euro) with a recess made with an angle grinder.
Yes, it's crude method but it works! I first practised on a bronze ring to get the hang of it, an then the real thing. It needed some smoothening but in the end the first item of tis build was completed.

 

Hi Jos,

Very interesting project,

Quiet an enterprise building this engine without a mill,you have my respect.

I'm currently about a month in a radial build myself so thats another dutch radial build.

I'm having an attempt on building the 7 cylinder radial from Volker Jung,all metric plans.
After contemplating on the edwards 5 I discarded them being to complicated to scale up or down to converting them to metric.

Later this week I will start a build log of my attempt on the 7 cylinder radial on this forum.

There is currently a guy on the dutch modelbouwforum who is in the completion fase of building the edwards five with several modifications.

I wish you much succes in this build and I'm going to follow all your progress.

groeten uit Eindhoven

 

Off topic, but hope to see your build pics & also providing link of same engine in case you hadn't seen it. http://philsradial.blogspot.ca/

 

Thanks Pat1311, I 'll check out the modelbouwforum. Converting to metric is quite a job but I took a year to do it. If you do it bit by bit it is'nt that hard , ideal for those moments too short to get into the workshop......

Jos

 

Sorry offtopic again...

Thank you for the info.

I've read his blog and his engine is really a beauty,he's a bit of a CNC wizzard too.
Only I think he's building the nine cylinder version.

 

sorry meant to say same engine designer

 

So, with the camring finished at last, it was time to take on the bigger parts.
As I mentioned earlier, I re-designed the crankcase so it would be made in two pieces. There are different reasons do do so.First of all, the material I had is a hollow bushing,being a rejected part of the recoil shock absorber of a M109 howitzer I took home when I was working as a QA engineer at a company which overhauled military equipment some 25 years ago. Also, it was easier to machine the bearing fittings for the crankshaft and idler bearings on a separate
part. And, if I mess up the bearing housings it would not be necessary to make a new crankcase, only a new bearing cover.
A year ago I made a fixture for machining the flat sides of the crankcase . I practised on a wooden model and now I could find out if it worked out on metal as well, which it did. I did not know what kind of aluminium is was but by machining it gave a magnificent bright gloss....great stuff, that.
After drilling with my drilling jig, I made the bearing cover and the idler assy. For fitting and checking the mesh of the gears, I used the dummy crankshaft I used in the wooden mockup and a dummy cam housing I made from a stack of perspex rings bolted together.

 

Now with the crankcase done is was time to get on the biggest chunk of metal in this build: the cam housing. Once again I used a rejected-M109-hydraulic-recoil-shock-absorber-part ( are you still with me?) .The diameter was some 12 mm. more than needed which was OK because I wanted to make the cam follower guides longer the the original design so that the cross loads on the cam followers were more evenly spread, and to reduce oil leaking. Downside on this was that it needed additional milling and ,in turn, the inevitable fixtures.

Turning the inside was pretty straightforward. The challenge was to drill the cam follower holes evenly spaced. After machining the inside I made two grooves marking the axial position of the cam follower holes.For the radial position I made a device consisting of a ring fitting exactly on the mating surface with the crankcase. On this ring I made a disc fitted with a steel bracket. The disc could be fixed on the ring with a scew. On the ring I made a mark, and also on the cam housing. The edge of the bracket was set on 72 degrees and along this edge a second mark was made on the cam housing.
Working around the circumference of the cam housing I made five marks.If the last mark was not exactly in the same position as the startng mark I adjusted the bracket and started all over again until all marks were evenly spaced. After that, I made punch marks where the radial marks crossed the earlier machined grooves. Then is was just a matter of drilling and reaming, everything all pretty basic.

Next challende was milling the cam follower slots at the inside. First I needed a way to fix the cam housing on the cross-slide of the lathe. From the scrap box I found a piece of 90 degree consruction steel which I fitted on the cross-slide . To make sure that is was perfectly flat I made a cutting device from an old core of a speaker( again, from the scrap box) and a 90 degree cutter bolted on is. Using the lowest possible RPM of the Emco which is 100 RPM I planed the steel. After drilling and tapping, the cam housing could be mounted .Slot milling was carried out with a small disc mill fitted on a home made mandrel.

At this point, the crankcase came into the picture. Cam follower holes were aligned withe the cylinder holes and holes for fixing the crankcase to the cam housing were match drilled and tapped.

With the critical operations done, the rest of the machining consisted of removing excess material. First the outer diameters, leaving a ring where the cam followers are located. After that I used the milling ficture to roughly remove the material between the cam followers. For the finish machining I made a rotating feed device consisting of a spindle, crank, pickup bracket and hinge blocks. I made five M5 holes at the front of the cam housing to which a hinge block was attached . These holes disappeared after finish machining.

With all the machining done I think the part looks rather nice..

Jos

 

Ingeneous and excerlent workmanship using what you have and not what you
would like to have.Considering the level of expertice required for this project
and your obvious skill it amazes me that you have attemped such a difficult project without a mill.You have my admiration

 

Thanks Bazmak, I have been doing milling operations on a lathe for about forty years so once you get used to it thing become a lot easier. It only takes more work in making fixtures etc. but I like challenges like that.

Jos

 

With the big lumps, crankcase and camhousing, complete, it was time to focus on the core of the machine, the crankshaft.
Single throw crankshafts have been used for centuries in man-powered machines until James Watt used it for the first mechanical powered rotating machine.
The one-side supported crankshaft is basically a simple part. The big trick is to get the crankpin hole parallel to the shaft centerline. In order to make the work on the lathe I fitted a thick cast-iron plate ( which was a rejected pneumatic cylinder cover)together with a U-shaped alu profile on the cross-slide of the lathe.The thickness of the plate and the profile was exactly enough to fit without clearance underneath the pre-machined crankshaft between the centers of the lathe. After that, I clamped the crankshaft onto the base plate an so , after releasing the tailstock, I had a perfectly aligned crankshaft. I only had to move the cross-slide the required distance to obtain the correct stroke of the engine . Drilling and reaming took less than a quarter of an hour( much less than it took to prepare the baseplate and clamp) and the I had a perfect clankpin hole.
To determine the length of the various diameters I first made a dummy crankshaft of scrap aluminium.After that, it was a matter of turning, measuring, polishing maesuring etc. to obtain the correct bearing fit.
Milling the crankweb was done on the toolpost with the mill in the chuck.

After completing the crankshaft with propdriver and spinner I could not resist the temptation to fit everything on a stand, with the prop, to see what it looks like.

So far, so good.....


"Failure is always an option" Mythbusters

 

Next step, moving outward from the crankshaft there are master and link rods. Conrods are basically a piece of metal with parallel holes in it For the link rods, I clamped a piece of alu on my angle plate fitted on the cross-slide of my lathe and drilled/reamed the holes .Finishing was the way sculptors do: just remove the material you don't want...
The master rod was of course a bit more work. The centre hole was turnes to achieve the correct needle bearing fit. After that, the rest of the drilling/reaming was done- again- on the angle plate. To mill the slot for the link rods I fitted the master rod on the toolpost, shimmed to botain the correct height. Since the thickness of the master rod was a bit oversize, I could correct small deviations to ensure the slot was exactly in the middle.

Well, on to the next step: pistons, liners and barrels. Yet another challenge....


Failure is always an option Mythbusters

 

Next in this build I started with the cylinder heads. First the combustion chambers. Since the conical shape is hard to measure I made a gauge to check the correct depth. Afther that, the holes were drilled with the help of my thrusty drill jig.

After drilling, then came the tricky bit: plug and valve holes. Having no milling facilities, I made a fixture
to drill and turn the holes on the lathe, see the previous post. Befor the actual drilling , I made a small punchmark with a center drill to see if the setup was correct. After thet I could drill and turn the hole. This was bigger than the original Edwards design because I chose for valve cages. Because the depth of the valve cage chamber is hard to measure against the curved surfaace of the combustion chamber I made a plug gauge to check both depth and diameter of the chamber.
Now this build is 80% of the origininal design but one dimension remains unchanged: the glow plug hole.
To check if there is no unwanted interference between pug and valve cage holes I made a plastic model. Having done that, it seemed that there were no problems.
With all holes complete, I went on to the slots for the rocker arm brackets. Again, I made a fixture ( there is no end in making fixtures..) fitted on the toolpost. This served for both driilng of the mounting hole and milling of the slot.
Next, the head fins had to be milled. I made the fins slightly wider to provide more support surface for the rocker brackets.
I did this with the help of my low-cost DRO to get the required distance between the cooling fins. Milling with a circular millis no big deal, only is is very messy, is produces a kind of aluminium mud all over the place.
Following this, I made the flat mounting surfaces for in and exhaust flanges withe a rotating tool in the chuck.

 

I have heard the expression , don't use a lathe as a mill,
or your mill as a lathe. Obviously, an uninformed expression.
Very interesting to see your lathe as a mill set ups.
Very clever to use the preexisting gears.
Thanks for a very informative build.

 

With the top fins complete, I considered the use of glow plug caps. I purchased 90 degree caps which made it necessary to cut away a notch for the cap. After the messy head fin milling I gave the head a ultrasonic cleanup. It is amazing how much muck this process removes , it looked much brighter afterwards.

 

I have been thinking a lot about the use of valve cages in this design. The originla dsign without separate valve guides and seatings was no option for me. The valve guides sticking out on top of the cylinderhead were an obstacle for the milling of the cooling fins. On top of that, I used free-cutting aluminium which is not as wear-resistant as the recommended 7075-T6 alloy. With valve cages, however, there is only a narrow edge between the valve seat and the port,see sketch. So I will drill the ports a bit higher to provide more sealing surface between cage and cylinder head.

Then there was the fitting issue. Exhaust valves have a hard life. After they are heated up during the exhaust stroke they have to get rid of the heat. This is partly achieved by cooling of the fresh mixture during intake and compression and by transferring heat to the valve seat when the valve is closed. The valve seat in turn must transfer the heat to the material of the cylinderhead. I have been considering the use of epoxy to get a gas-tight seal between cage and head . This might work for short demo runs but I intend to use this engine in a R/C plane ( if it delivers enough power....) which will cause a considerable thermal load on exhaust valves and seatings. So I think a press fit will be the best choice. n interfenence fit of 0,02 mm seemde a good idea so I made a test cage and a piece of alu bar stock in which I made a chamber just like in the cylinder head. With some pressing tools I made the fitting went well so I went on with the 0,02 interference in mind.
For the cage material there were several options. There is , of course, cast iron, phosphorous bronze, and aluminium bronze. Phosporous bronze, so I have heard, is difficult to machine, CI seemed a bit brittle to withstand the pressing forces. Aluminium bronze, however, is easy to machine and is also used for valve guides in "real"engines. I manages to get hold on a length of 1 m 10 mm dia. I t didn't come cheap, 50 euros incl VAT but now I have enough material for this and possible future projects...
Because not all valve cage chambers were exactly the same I had to measure them, record the diameter and make all cages to size for each induvidual cage chamber.
So I ended up with this. Next phase: press fitting them.

To be continued..

I did not fail.. I just found 10.000 things that didn't work Thomas Edison