My Hodgson 9 Radial Final Assembly

[mayhugh1]

Here I've installed the rear main bearing, crankshaft, the oil pumps, the rear seal, impeller, and the distributor drive gear. This was just to verify the fit that I had when I machined my crank. During trial fitting of the crank it came in and out of the crankcase many times as my goal was to center the crank for a max .002" without the use of shims and to also set near-zero backlash with the distributor also without shims. - Terry

 

[mayhugh1]

Here is a photo of my setup to time the camshaft. I made a generic degree wheel and positioned my height gauge to act as a pointer for it. I made a fixture to support a plunger style DTI in the spark plug hole in order to indicate TDC. A second DTI rests against the intake and/or exhaust lifters to probe for the centerlines of the cam lobes. My technique was to probe for the .015" cam lobe heights and then divide by two to find the center. I was careful to always approach the measurements with the crank turning CCW to avoid backlash errors. The goal here is to move the cam ring in relation to the crank in order to center TDC between the intake and exhaust lobes. This is done by carefully lifting the cam ring jackshaft up and out of engagement with the crank shaft and then rotating the cam ring for the best result. Trial and error is at work here. Unfortunately, with the gears used in this design it turns out that the resolution in terms of cam degrees is at best 2.8 degrees per cam ring tooth. This is reflected over in crank degrees by a factor of 8 giving a whopping 22.4 degrees resolution in crank degrees. Fortunately, the phasing of the two gears on the jackshaft, 32 tooth and 12 tooth, can be used to advantage to cut down this resolution error. This means the jackshaft has to be disengaged from both gears and then rotated and then the lobe center is re-found as described above - a bit more complicated trial and error. I called it quits when I had TDC centered between the lobe centers by +/-4 crank degrees. This was equivalent to the resolution to which the cam ring was designed. The cam timing I measured is a pretty common result for a mild street engine. The exhaust valve opens 43 deg before BDC and closes 7 deg ATDC. The intake valve opens 2 deg before TDC and closes 35 deg ABDC. The intake duration is 208 deg and the exhaust duration is 221 deg. The distance between lobe centers is 222 deg. I've done cam timing on full size engines before, but the hardest this thing for me to wrap my head around on this step was remembering that the crank on this engine is turning 8 times faster thatn of the cam instead of 2 times faster. The distributor timing still needs to be set up but that will be one of the last steps in my assembly. - Terry

 

[mayhugh1]

Here is a photo of the cam timed and finally installed. Also, the slave rods are installed. As it turned out I thought I could install the slave rods with the front main bearing installed, but that wasn't possible. And so I had to do a bit of disassembly to install them. Hopefully the crank will not have to come back out. Even with the clamp arrangement I used to tie the front and rear sections of the crankshaft together, the fit is so close that it is difficult to pull the front section out even with the clamp screw removed. The second photo is a trial fit of the front cover and I couldn't resist adding the prop and spinner temporarily. As mentioned before, the spinner is polished aluminum with a stainless hex pressed in the rear. The stainless insert is tapped for a reverse thread so the prop won't loosen while the engine is being started. If you look closely, you might notice I engraved a warning about the LH thread on the side of the stainless insert. I always do this on left handed parts as a courtesy to the next person who might need to loosen the part. Hodgson design vents the crankcase through a hole drilled through the center of the crankshaft and a second intersecting holed cross-drilled hole in the crankshaft inside the crankcase. I would have guessed that the oil scavenging pump would have done away with the need to vent the crankcase since it would be sucking in this area and venting into the oil tank which is also vented. But just in case, I kept his ventilation system. Since his design was expecting a simple nut on the end of the crankshaft, I had to provide a path for the venting through my spinner. That is what the radial holes in the hex portion of the spinner are for. The diameter of the portion of the crankshaft extending through the front cover sure looks small for that big 23" prop. Another thing about radials is that our model engines typically don't drive loads and so really don't develop much actual horsepower unless connected to a dyno or some other load. A radial, on the other hand, is turning a propeller and trying to fly. - Terry

 

[aonemarine]

People like to give thier engines names, and if I were to name your radial I would call it "dedication". That is some beautiful work. Maybe some day I will make it to your level of workmanship.

 

[mayhugh1]

Chuck,
Here are some construction photos I took while I was building my distributor. The slot in the bottom of the metal housing is where the Hall device is placed. There is another slot on the opposite side of the housing. Only one is visible in this photo. The Hall device leads come out this slot at the bottom of the distributor. The white Delrin disk in the unassembled parts photo - the one with the large center hole and small perimeter hole - goes into the metal housing on top of the Hall device and sandwiches it between the aluminum housing and this disk. The dark shadow on this white disk directly opposite the perimeter hole is a milled cavity for the Hall device. The small perimeter hole is actually tapped for a 2-56 screw. A screw comes up through the other slot in the bottom of the housing and secures the disk and Hall device in place. When loosened, this screw can be used as a handle to slightly adjust the angular position of the Hall device in order to tweek the timing. This is necessary because the resolution of the positioning of the rotor and the number 1 high voltage tower is set by the finite number of teeth on the distributor gear. - Terry

 

[tony_m_baker]

Do you have any drawings for this engine? I am really only interested in the parts show here. I want to do something similar in a sculpture I am doing and the radial engine is a perfect model.

 

[mayhugh1]

I thought I might comment on the valves in this engine. I really like Hodgson's valve design with its beefy .187" stem diameter necked down only behind the head where it is needed to unshroud the valve and increase flow. The valves on the other two engines I have built had tiny .125" stem diameters and were difficult and tedious to machine. Hodgson's valves turned out to be a breeze to machine.
I'm fortunate enough to have a 9x20 cnc converted lathe. It has its issues, but I've learned to work around most of them; and with the right amount of babysitting I can hold .001" or so for a while. What I did was to saw off 15 half-inch diameter 303 stainless blanks which were 4-1/2 inches long. I generated 2 cnc programs - one to rough out the valve but to leave .020" stock along the entire profile of the valve. I used somewhat aggressive feeds and speeds because I wasn't interested in surface finish at this point. And, because I was leaving a good bit of stock I didn't have to continually monitor the tool calibration in order to hold cutting accuracy. I ran this program 30 times - once on each end of each blank with the same cutting tool. I was able to rough out 30 valves in 3 hours or so by just zonking out and feeding blanks to the machine. My plan was to then to stress relieve them overnight at 450F in my oven before running the finishing program. Well, we took a one week vacation between the roughing and finishing steps and I forgot to do this step. I'm not sure if it is really necessary but Bob Shores recommends it and I figure it can't hurt.
Calibrating my lathe's x-axis for sub .001" accuracy is touch and go and when cutting multiple parts like this I don't like to change tools because it adds another set of variables to the accuracy of the cutting. From past experiments I have found a Kennametal DCMT2151UF light finishing insert can give a great surface finish on 303 in my lathe even when cutting as little as .002" off the diameter. The DCMT insert is required for this valve to get access to the entire profile of the valve. Again, after calibrating the x axis and touching off z, I just feed parts into the machine. Only this time, after every run, I measure the diameter of the valve stem in the area where it will make actual contact with my valve guide. (This is the only part of the valve whose diameter is really critical.) After every run I reset the x-axis calibration to agree with the measured diameter of the stem in preparation for the next run. Much less agressive feeds and speeds more than triples the finishing time, but that gives me more time to surf the web while the lathe does the tedious work. I was able to finish all the valves in an evening on a single insert and the wear on the insert was negligible. I normally aim for .0005" valve stem clearance with the guide and this is what I was able to more or less achieve. I programmed a very slight taper at the top of the valve stem so there would be no issues with inserting it through the valve guide. I later had second thoughts about whether or not this engine depends on a larger valve stem clearance (Hodgson spec'd .005") in order for the guide to get some blow-by for lubrication in addition to the slop added to help the valve find its own sealing position. There is no top end lubrication except for oil that makes it way into the combustion chamber and out past the valve. So, later I came back and in a secondary operation I polished another .0005" off the completed valve stem. After the finishing program was run, I sawed the valves off each end of the blanks and then performed all the tedious individual operations to face the valve to length and to add a keeper groove. The valve retaining scheme was my only departure from Hodgson's design (plus a small reduction in diameter for my valve cage). He uses a tiny pin in a cross-drilled hole in the stem to retain the valve. Instead, I machined c-washers that slip into a groove I cut into the top end of the valve stem and then retaining cups to secure the c-washers. It's a lot neater looking, but is one of those things that no one notices since they are hidden out of sight under the rocker arm supports. - Terry

 

[mayhugh1]

Tony,
Drawings are available from the original designer at

agelessengines.com


Regards,
Terry

 

[kuhncw]

Terry, thanks for the information on your distributor. That is a nice design.

Chuck

 

[swilliams]

Awesome Terry, really top notch workmanship. I'll get back and read all of this carefully when I have some time

Cheers Steve

 

[mayhugh1]

Here, I've installed the rings on the pistons and, miraculously, didn't break a single one. I ended up with lots of spares of compression rings but only 1 extra oil ring. Hodgson's oil ring design is fairly complex with a shallow circumferential groove cut into the outside of the ring and a number of tiny radial drilled holes. The theory is that oil is scraped into this groove between the ring and cylinder wall, forced through the holes in the ring, and then through a second set of radial holes cut into the piston where the oil can return to the crankcase. Hodgson evidently changed the piston/ring design in the later versions of this engine because I've seen others that had only a single compression ring. The drawings I have call for two compression rings. They are cast iron and only .045" wide and so they look pretty fragile compared with the ones I made for my V-4. I had to make all new fixtures to fabricate and heat treat these rings because they were a different diameter from those on the the V-4.
One thing about this project is that fixtures are a way of life. I think I ended up building, on the average, one unique fixture for every three unique parts built for the engine; and I didn't even build all the recommended ones. This weekend I will install the cylinders and the pistons on the crankcase. I have marked each cylinder and piston with a unique number so I can keep track of them should I need to later disassemble the engine. I have three spare cylinders and ringed pistons. Their measurements, leakdown test, etc are all recorded in a notebook to help me keep track of what works and what doesn't work later when/if the engine is running. This step will be tedious and take a bit of time as there are 72 small pattern 4-40 nuts that have to be tightened a fraction of a turn at a time to secure the cylinders. I also need to add a washer and lock washer on each stud with my fat fingers without dropping anything into the engine. I cut out some .004" linen-paper gaskets on my Tormach using their vinyl cutter to help seal oil leaks between the cylinder and crank case. This will reduce the compression ratio even more. I'll take a break from nut tightening this weekend to include a section on compression ratio calculations for this engine. - Terry

 

[petertha]

Instead, I turned the whole crank from a single scrap round of mild steel and drilled the hole for the crank pin before sawing the result into two pieces with guaranteed alignment. The cheeks were slit so the master rod journal could be clamped into place with an 8-32 SHCS in each cheek.

I'm really interested your alternate procedure of constructing the crank shaft. Do these sketches roughly represent the machining sequence up until slitting part? By 'guaranteed alignment' you mean drilling the crank pin hole through the entire segment of what will be become front & rear counterweight pieces when sawn apart? What is the crankpin itself made of? Do the ends have a reduced diameter step which engages the web holes?

 

[mayhugh1]

Petertha,
You have it exactly right in your drawings. My crankpin is hardened O2 tool steel (and I added a bronze bearing to the main rod). My crankshaft is 12L14. I drilled and reamed the hole through the cranshaft blank for the pin after turning the ends but before sawing the crank into two pieces. In my radial engine I turned the crankpin only a few tenths below the diameter of the hole for a snug sliding fit. This was in case I might have to later disassemble the crankshaft within my crankcase due to some problem. This clearance would make it a bit easier to separate the two pieces within the crankcase. (Remember, in this radial the master rod captures the crankshaft within the crankcase.) In my radial, I locked both cheeks to both sides of my crank. I was able to do this because my main bearing alignment ended up pretty good and I had no binding. I don't think it is necessary to lock both of them. One is sufficient to keep the oil passages aligned and to keep the pin in place. In my V-4 I only locked the rear section. I reamed the crank cheek in the front section in this engine .001" over-size and did not clamp it. I did this to relieve a small crankshaft bind due to a small misialignment in the main bearings in this engine. This engine had 5 main bearings and a center section high pressure Babitt oil seal, and my machining just skills weren't up to the task of getting them all perfectly aligned. -Terry

 

[mayhugh1]

I finally got all the cylinders installed on the crankcase. It wasn't as tedious as I thought it would be, but I did have to make a special wrench to tighten the nuts on the two lower cylinders on either side of the oil sump. After 1-1/2 years of accumulating parts in plastic bags, this engine is finally starting to look like one. I trial fitted a pair of my rocker covers to check the length of the pushrods that I made. I'm going to shorten them by .050" to get the valve adjusters in a better looking position. I'm also having second thoughts about my rocker covers. I'm not sure I like their looks as much as I thought I would. I might redo these.
I have now put the engine into an ugly but effective setup to "motor it in." I'm using an electric drill and my starter adapter to spin the crankcase to check for any last minute problems before I ask it to produce its own power. I want see the oiling system in action and check for leaks that are easier to fix now. This will be my first real test of my drip-feed oil tank as well as the scavenging oil pump's ability to evacuate the sump. I was able to check the pressurized oil side earlier when I finished installing the crankshaft. I've opened the closed loop oiling system for this test so I'm not recirculating dirty oil during this test. I'm just collecting the oil from the scavenger pump for disposal and refilling the oil tank as needed. My plan is to run about a quart of oil through the engine in a number of start and stop runs at various speeds. I'm not expecting this step to seat the rings without the forces of combustion pushing the rings against the cylinder walls. It should start the process, though, by wearing off any high spot imperfections in my rings. I may be able to see this debris collected on the magnet that I embedded in the sump cover.
I'm amazed at how much oil is being moved through this engine. Even at 1 rpm the outlet hose of the scavenger pump is evacuating oil at a rate that I would have expected at idle. I now see the importance of adding the oil drip feed system as recommended by Hodgson. I've left the screws out of the front cover so I can watch for oil build up in the front portion of the engine which is a known issue with this design. I enlarged the area of the drain-back system some 50% and I'm curious to see if this at least partially cures it. The collected scavenger oil is pretty black, but I think this may be the bluing inside the cylinder being scrapped off by the rings. - Terry

 

[mayhugh1]

While 'motoring-in' the engine I decided to try to quantatively check the compression in each of the cylinders. For this engine build, I made the compression tester in the photo. I designed it for use with a model engine with negligible volume below its check valve. The chrome center portion of the gauge was robbed from a $5.00 peak reading tire tester from a local auto store. I added a spring-loaded ball check valve in the bottom portion which is threaded to screw into a 10mm spark plug hole. At the top I adapted a 0-100 psi gauge that I bought at the NAMES show last year. Since the pushrods and rockers aren't yet installed, I have to manually pump one of the valves with my finger about once per crank revolution to allow fresh air into the chamber while the engine is spinning. After a half dozen or so revolutions, the peak compression pressure is accumulated by my peak reading gauge, and the reading should approach what I would obtain from a conventional compression test.
After passing 125 mL of oil through the engine during 'motoring,' I obtained the following measurements:
#1 = 83 psi
#2 = 68 psi #9 = 83 psi
#3 = 72 psi #8 = 72 psi
#4 = 72 psi #7 = 72 psi
(I did not have enough clearance in this setup to screw my gauge into cylinder numbers 5 or 6.)
Taking an average reading of 75 psi, this would make the average dynamic compression ratio = 75psi/14.7psi = 5.1
The compression ratio spec'd by Hodgson for this engine is 6.7. I computed the static c.r. for this engine using Hodgson's drawings since measurements of all my parts that affect the c.r. agree almost exactly with his drawings.
Here's what I obtain:
Head volume due to conical section of head = (pi/3)(r^2)h = (pi/3)(.6025^2)(.281) = .107 cu.in.

Head cylindrical volume due to thread relief = (pi)(r^2)h = (pi)(.6025^2)(.030) = .034 cu.in.

Head cylindrical volume due to aluminum washer offset = (pi)(r^2)h = (pi)(.6025^2)(.030) = .034 cu.in.

Cylindrical volume from piston top below top of cylinder at TDC = (pi)(r^2)h = (pi)(.5^2)(.062) = .049 cu.in.

Total volume at TDC = .107 + .034 + .034 + .049 = .224 cu.in.

Volume at BDC = .224 + (pi)(r^2)(stroke) = .224 + (pi)(.5^2)(1.125) = .224 + .884 = 1.108 cu.in.

Static c.r. = 1.108/.224 = 4.9
I believe Hodgson left out the excess volumes due to the thread relief and the aluminum washer used as the head gasket in his calculations because the result is 6.7 if these are ignored.
I plan to make a final compression test after the motoring is finished and the pushrods are installed to see if the extra motoring time improves any of the pressure readings.
Terry

 

[swilliams]

Nice Terry. Will be interesting to see what happens after more motoring. With a little luck the rings may bed in making things more consistent between all cylinders. So are you using the 10mm spark plugs as sold on the Jerry Howell site?

Cheers Steve

 

[Lakc]

I didn't have time to jump in on the mailing list when this subject came up, but I have a few moments now to comment.
There are a lot of factors that go into a compression test that makes the values obtained less then useful for determining compression ratio.

Leakage is probably the biggest one, and that is what makes it a useful test in servicing engines when something goes wrong. It will take actual run time to help the rings and valves seal and you should see compression rise slightly after its run. Oil itself will mask sealing issues at cranking speeds, so you end up with a spread based on where your oil has sprayed/dripped/collected. Leakage in your compression tester from your check valve may not be as repeatable as you would like.

Throttling and pumping losses due to restrictions and pressure variations in the intake and carburation. Air is a fluid, and any non-laminar flow involved will vary the amount of intake each stroke. At higher running rpm's this air has mass and momentum that will tend to lessen those effects. Although not applicable to round engines as far as I can tell, some V engine designs can have two cylinders trying to draw air simultaneously from relatively confined spaces which lead to less intake. In our "minimal area" type model compression gauges this adds even more reduction of the "true" pressure in the cylinder

Cranking speed variations will also vary your compression readings. Unless your driving a crank with a very large mass at an exact speed every time the work of compression will slow the crankshaft a different amount.

So, in a perfect world the math would work out, but its not going to cooperate that simply. To do a proper compression ratio check you use liquid to measure the volume of surfaces involved. Much less variables to affect your readings that way. Dont sweat the differences too much either, due to manufacturing tolerances stacking up the compression can vary by a significant amount. One auto manufacturer listed a delta spread of compression ratios for its V10 engine as 8.9-10.1:1.

 

[mayhugh1]

Jeff,
Points well taken.

Steve,
I bought my CM-6 plugs from a vendor at Names last year. Cncengines.com sells them for $5.00 which is as cheap as I've seen them. They're really big and way out of scale for this engine, but they have better resistance to oil fouling and this will probably be welcomed for cylinders 5 and 6 at the bottom of the engine. If I'm lucky and my engine doesn't have a lot of oil fouling issues after it is running and dialed in, a possible option is to make some adapter inserts for 1/4-32 plugs which will fit the scale a lot better. - Terry

 

[mayhugh1]

Petertha,
Here are some photos I took during the construction of my crankshaft. - Terry

 

[kvom]

So how did you drill the hole so as ensure it would be precisely parallel to the shaft?