My Hodgson 9 Radial Final Assembly

[mayhugh1]

Kvom,
Sorry I didn't have a photo of that setup. I held the crankshaft by the front turned bearing in a 5C collet holder, and indicated it vertical in my mill vise. The spotting, pre-drilling (carbide drill), drilling, and 2 step reaming was done on my mill all in the same set-up. - Terry

 

[vihatch]

Thank you for a very enjoyable thread. I am currently 95% finished with a Howell V twin. Your comments about valve testing and lapping coincide exactly with my experience. It's good to know I am not the only one who doesn't get a perfect valve seal. I am going to try the 1200 grit lapping compound on my next build.

 

[seagar]

wow!!!!!

 

[mayhugh1]

These are photos of the assembly stand that I built for the Hodgson. It is really necessary to have something like this to hold the engine while it is being assembled. The engine is awkward to hold and very heavy, and it must be held in some a number of positions while several fussy assembly steps are performed. This is no time to drop the thing on the floor. Some have even built a special rotating engine stand for the engine. I decided on the simple design in the photos which lets me get to most areas during assembly and, so the work wasn't throw-away, I integrated it into the final running display of the engine. The support ring that bolts the stand to the rear cover uses the rear cover mounting holes to mount the engine to the stand, but it leaves three of them free so that the distributor timing isn't disturbed when the engine is taken on and off the stand. The stand is designed to temporarily support the engine from its front also so that the distributor/rear cover section can be taken on and off easily when it comes time to time the distributor. - Terry

 

[mayhugh1]

I've called it quits on the "motoring-in step." I ran about 250 ml oil thru it in a large number of short runs over several days while I take care of some other things. I had to re-charge the batteries in my 18V drill three times in the process. I took a photo of the sump magnet after moving about 200 mL of oil through the engine which took about 30 minutes total run rime. There is a very fine layer of metal over a portion of its surface which is probably coming from the rings. I had the oil drip feed system set at about two drops of oil per second for this test. This seemed to keep oil moving through the scavenger at the slowest rpms my drill was turning over the crank. At any fixed drip setting the actual flow rate varies with engine speed and so I expect this is going to be a critical setting with the engine running. I never saw the scavenger pump lose its prime in the 3-4 days this test ran. Behind the front cover the entire cam is wet with oil and so the front portion of the engine is getting well lubricated even it low rpms. The is no oil leaking past my rear seal into the fuel diffusing area and this is good news. The best news though is that the engine does not seem to be pumping significant oil into the combustion chambers of the cylinders. There is nothing leaking out the spark plug holes of the bottom cylinders and looking with a flashlight into the other cylinders shows only a few drops of oil in the combustion chamber. This was not the case during a similar test on my V-4. The engine pumped quite a bit of oil out the cylinder decks even with the crankcase well vented. I may finally have made a decent set of rings. I had to plug the lower lifter bushings with silicone plugs to keep oil from pouring out them during this test. When I finally install the lifters I'll select the ones with the closest fit to the bushings at these low positions. Leakage at these lifters is a known problem with this engine design and is the reason why I milled the oil-catching trough in the bottom of my assembly stand. The next step is to install the rockers, intake and exhaust pipes and to time the distributor. Al though totally unrelated to this step in the assembly, I'm also including some photos of the construction process during the past 18 months. - Terry

 

[Lakc]

Is that the newer design diffuser or the older one?

 

[mayhugh1]

Lakc,
I believe this is the latest one. The planset supplement that I bought 2 years ago contained this large diffuser. - Terry

 

[danthompson58]

Very nice build, excellent workmanship and pictures with descriptions.
Dan

 

[mayhugh1]

The next step is to time the distributor and button up the rear of the engine including the rear cover and air guide. The "motoring-in" test showed the oiling system is working well with no engine oil leaking into this area, and so now it should be safe to assemble the rear of the engine for the last time.
There are two separate timings involved with the distributor. In the first, the trigger signal of the Hall device with must occur when the rotor is directly below one of the high voltage terminals in the distributor cap. I established this in my distributor when it was finally assembled and tested about 9 months ago. In my design, the Hall device is buried in the bottom of the aluminum distributor housing and attached to a Delrin plate that can be rotated +/- 30 degrees with respect to the distributor cap and then locked permanently in place. This has already been done and tested.
The second timing requirement is to time the Hall device to TDC on the crankshaft comprerssion/firing stroke of cylinder #1. The only way to do this in this engine is by trial and error by repeatedly mating the rear section containing the distributor onto the main crankcase section to engage the distributor pinion gear with the crankshaft pinion and then checking that the Hall device triggers at TDC of cylinder #1. The rear section is then removed, the distributor pinion gear rotated one tooth, and the rear section is re-mated onto the crankcase until the best result is obtained. I included a stationary timing pointer and a set of 10 degree (crank) timing marks on either side of TDC on my distributor housing. So, I located the best pinion gear engagement that gave me a Hall device trigger at TDC as indicated by my dial indicator in the sparkplug hole of cylinder #1 simultaneously with my TDC timing pointer on the distributor housing. In addition, since I engraved cylinder numbers on my distributor cap beside each HV tower the rotor also has to be beneath the tower maked #1. (I added these distributor cap numbers in order to make it easier to keep track of my plug wires later.) Once this is done, the timing can be advanced as needed for best engine performance by loosening a locking screw on the distributor housing and rotating it. This will advance or retard the Hall trigger with respect to TDC on the crank but not disturb the relationship of the trigger with respect to the HV towers in the cap. In addition, it will rotate the calibrated timing marks on the side of the distributor housing with respect to the stationary timing pointer so the amount of the advance can be read. There are two blue leds - one of either side of my firewall - which turn ON when my Hall device is ON. The length of time the Hall device is ON is the dwell time. This is the time during which the ignition coil charges. For my TIM ignition the dwell time is fixed and is established by the radius of the mounting circle of the nine magnets (and their diameters) that make up my trigger disk. For my radial engine I've set the dwell at about 22 crankshaft degrees. The dwell of my V4 was 25 crankshaft degrees and this was sufficient for a useable spark at 1700 -6200 (measured) rpm without excessive overheating of my coil or output transistor. I decided to maintain about the same dwell for my radial since I will be firing 5 more cylinders and the additional power dissipation could be a problem. I don't plan to rev this engine much above 3000 rpm anyway as I expect that will be scary enough with a big 23" prop. Of course my choice of dwell is very dependent upon the inductance and resistance of the coil being used as well as the maximum charging current of my output tansistor. Since the 'old' style Exciter coils that I have been using are no longer available, this will have to be re-thought for any future engines.
In my other two engines I have found these dwell leds to be invaluable when setting up and troubleshooting the ignition. The secondary or high voltage portion of my TIM ignition is enabled by a separate switch and so it is not necessary to attach a dummy sparkplug or sparkgap when checking timing or troubleshooting the low voltage section of the ignition. Once the distributor is timed it is attached for, hopefully, the final time and secured with three SHCS at the rear of the engine. The assembly/display stand clears these screws so the timing will not be disturbed when the engine is taken off or attached to the stand. I am attaching a SolidWorks cross-section of my distributor. The actual machined components shown in an earlier post can be located in this drawing. - Terry

 

[mayhugh1]

I'm now installing the intake and exhaust pipes. In the photo you can also see the wire loom that I've added to help organize the plug wiring that will encircle the engine behind the heads. I've also added a photo showing the fixture I used to silver solder the stainless steel pipes to the beefy flanges I made to fit the rear of the heads. This was the first silver soldering that I've ever done of any consequence and I was surprised at how easy it really was. I used 3/8" annealed 304 stainless tubing and it bent very easily and cleanly using a Rigid 600 series hand bender without heating or filling the tubing. This is a 3 roller bender that I was preparing to build myself, but Santa ended up bringing it for Christmas instead. I made the wood gauges to check the tubing bends before they went into the soldering fixture. The intake/exhaust manifold gaskets were cut from 1/64" automotive gasket material on my Tormach using their new vinyl cutter. I created a cutting program for the gaskets using my CAD model of the flange itself to guarantee a perfect fit. I was having so much fun learning to solder the pipes that I made almost twice as many as I needed. Everyone of them dropped into place with no interference. - Terry

 

[johnl]

Great work on your 9 cylinder. I recently received my set of plans from Lee H. and am gathering materials.
I hope mine turns out as well as yours did. John

 

[mayhugh1]

Here are some photos I took during the machining of the crankcase. There are 137 holes that must be drilled and tapped - most of them are 4-40's. If I had been more brave I would have cnc tapped them on my Tormach, but I chickened out at the last minute and decided to tap them all by hand. I'm proud to say I didn't break a single tap. One of the photos contains the fixtures that I had to make to support the crankcase in my rotary while I machined the crankcase. This photo also contains the first crankcase that I built and screwed up during its last maching step. If you look closely, you can see that the cylinder mounting bolt hole pattern is off-center from the cylinder deck. I had the tailstock tightened up ridiculously tight against the crankcase while it was on the horizontal rotary for this machining step. This caused the rotary to lose steps as it rotated the crankcase under the spindle and caused the error. I thought about trying to salvage it by plugging the holes and re-drilling, but I hate to start out a project on a questionable foundation. - Terry

 

[mayhugh1]

Here are photos of some more of the fixtures that I constructed to build the Hodgson 9 radial. This first group of fixtures were used to make pistons and rings. The heat-colored parts go together to form the annealing fixture used to heat treat the cast iron rings. The tool in the back is a cleaver for cleanly breaking the rings for their gaps. The wrench-tool in the second photo was used to screw the heads onto their cylinders with a .030" dead soft aluminum head gasket between them. This tool was also used to hold the completed assembly in my mill vise in order to locate and drill the mounting holes for the cylinders to the crankcase. Once this drilling is done the cylinder is married to that particular location on the crankcase and likely cannot be salvaged after disassembled for repair. In my case that represents about 100 hour chunk of time that would be discarded if, for instance, a valve seat would need to be repaired. This is why I made three complete spare assemblies. The third photo is a number of tools I used to fabricate and install my valve cages. - Terry

 

[metalmad]

Top job Terry
Love your work ;D
Pete

 

[petertha]

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.
Terry

Very interesting Terry. FWIW, I went through a similar calculation exercise on my own cad project. Another potential source of head 'volume add' that further reduces CR is the valve bottom superimposed into the cone shaped combustion volume. Depending on dimensional variables like how deep the seat/cage sits in the head, valve shape, valve seating depth, head cone angle etc... the resultant true head volume also develops those extra little stubby 'antlers'. It wouldn't seem like much when you see a typical drawing section view cut through the valve center & the valve bottom appears to connect closely across the head cone line. Its easier to visualize in 3D.

In my case it wasn't much. But a steeper cone angle + larger diameter valves + deeper seating... potentially adds up a little here, a little there. Mostly I was trying to get a feel for how much CR would vary just by expected newbie machining & assembly inaccuracies. My planned remedy was utilizing different thickness head shims to balance the cylinders CR, I guess after doing accurate liquid volume measurement. But I'm not there yet so cant tell you how it turned out.

 

[mayhugh1]

Petertha,
I agree with you. This number is very sensitive to machining errors especially when talking about c.r.'s of 8 or 9. In this radial I actually have the opposite case to the one you are looking at. My seats aren't fully recessed and so the valves actually rob some volume in the combustion chamber and will tend to raise the c.r. a bit. I didn't think to calculate its contribution when I did my c.r. calculations. - Terry

 

[mayhugh1]

Here, I've installed the intake and exhaust pipes, the pushrods, and the rocker arm supports. The engine is now completely assembled except for the carburetor and the sparkplugs and plug wiring. I shortened all my pushrods a bit to get the lash adjusters in the center of their adjustment range. I counted a total of 19 different machining operations that affect the length of this rod. My push rods end up about .130" longer than the stock rods due to changes I made in the valve design in order to accommodate my valve cages and keeper design as well as the changes to the lash adjuster that I made. The pushrods act at compound angles between the lifters and the rocker arms. These angles are different for the intake and the exhaust rockers. I studied the geometry of their movements and designed spherical cavities with conical walls for clearance in both the rockers and the lifters. The stainless steel pushrods have turned hemispherical ends to match these cavities. I hardened the lifters, and before milling the cavities into the aluminum rocker arms I pressed in phosphor bronze inserts. The rocker arms also rotate on bronze bearings. I was not able to convince SolidWorks, the CAD modelling program I'm using, to mate my pushrod with my lifters and rockers in an assembly so I could easily design the 'ball-and socket' connections. This was all done manually by trial and error and prevented me from getting a complete animated assembly of the entire valvetrain. I figured it was because the version I'm using is 6 years old but I was recently able to speak to a SW rep at a local demo and learned that this is a limitation of the tool. This is why I decided to make the pushrods a little longer and then trim them at final assembly. I made compression tests in each cylinder before putting the engine back into my 'motoring-in' setup.
The compression test results were:
#1 = 70psi #9 = 72psi
#2 = 63psi #8 = 72psi
#3 = 66psi #7 = 72psi
#4 = 66psi #6 = 80psi
#5 = 77psi
These readings are repeatable and reasonably consistent especially since the engine hasn't yet been run and the rings have not been seated. The maximum I should theoretically expect for this engine is 72psi. The two cylinders at the bottom of the engine give somewhat higher than expected readings. This may be because oil has drained into them (known problem with radials) and reduced the volume of the combustion chamber somewhat. These readings are not consistent which those I made earlier before the pushrods were installed and when I manually pumped one of the valves to get air into the combustion chamber to get a compression reading. The reason for this is probably one of those mentioned by Lakc in his comments to my first compression measurements. It is certainly likely that I didn't open the valve manually the same amount that the cam is now opening it.
I also connected a sparkplug to my coil to exercise the ignition with multiple firings while the drill is turning the engine. So far, I am getting nice fat sparks at the test plug that seemed to be properly timed with the position of the rotor under the HV towers. The fat sparks are expected at low rpms since the coil is having a lot of time to saturate. More on that later.
I'll run another 200 ml or so of oil through the engine with another series of short runs driven by my drill and starter adapter in order to start the run-in of the cam and lifters. I'm also curious to see what the oil seepage around the lifters on the lower cylinders is going to look like. I hate making engines that leak oil but I knew the reputation of this engine before I started. I'm finally happy with the appearance of the rocker covers now that they are finally all installed and don't plan on re-designing them. - Terry

 

[petertha]

The compression test results were...
These readings are repeatable and reasonably consistent...
The maximum I should theoretically expect for this engine is 72psi.
The two cylinders at the bottom of the engine give somewhat higher...
Terry

I am really enjoying your detailed build/assembly report!

Is the Hodgson master rod compensated to acheive equal CR's for its link rod pin layout? (as opposed to 360 deg / 9 cyl = 40 deg equal nominal radial spacing). I'm sure he thought of that, but was just curious because of the pressure readings. Does your 72 psi theoretical max come from CR calculatiion using your TDC/BDC volumes? (Pmax = CR * 14.7 psi atmospheric)

link & post 4&5 has related info FWIW. This notion didnt even occur to me until I somehow stubled onto the topic.
http://www.homemodelenginemachinist.com/f26/master-rod-layout-15156/

 

[mayhugh1]

Petertha,
In my post I'm estimating my max pressure readings from my calculated compression ratio and assuming 14.7 psi for atmospheric pressure.
No, the master rod is not compensated. The slave rods are all at 40 degrees. I went through the calculations to see what would be required for compensation but found there was not enough room for eight compensated and reliable slave rods on the diameter master rod that would fit this crankcase. If there had only been five cylinders I could have done it. I briefly thought about arranging the ignition trigger magnets and high voltage towers on the distributor in irregular intervals, but at that point in time I didn't feel confident about drifting too far out of the box on my first radial. - Terry

 

[mayhugh1]

I have finished the 'motoring-in' step with the pushrods installed in order to get a bit of initial wear on the cam and lifters and to check that they are all getting lubed. The two bottom-most lifters in #5 and #6 continue to leak oil. I found the cause and there doesn't seem to be practical fix for it at this late date. What's happening is that oil is accumulated behind the front cover of the engine where it is drained through a trough (that I enlarged by 50%) into the sump where it is removed by the scavenger pump. The inlet tube for the scavenge pump is about 1/8" from the bottom of the sump and so theoretically there would be a constant 1/8" level of oil left in the sump. However, air is trapped in the sump with no place to go when the engine is stopped because the only vent for the sump is the oil line going to the scavenger pump and this, hopefully, remains full of oil to maintain the pump's prime. Therefore the oil behind the front cover cannot drain into the unvented sump and it leaks around the lower lifters. Drilling a hole in the sump to vent this air might result in a more serious leak when the engine is running at speed. If I had known about this earlier I would have added a third tube to the sump and vented it into the crankcase. Hodgson recommends draining the sump after running the engine and so this should reduce this leakage. Hopefully this won't cause the scavenger pump to lose prime which is the reason that I was planning to not drain it.
I've finished the plug wiring. It took me longer than it probably should have because I'm anal about the appearance of the wiring in my engines. The wire looms that I designed for each cylinder keeps the wires away from the exhaust pipes and helps organize a radial harness around the back of the engine. Beside giving a better appearance, it also keeps the plug wires at right angles and away from the Hall effect sensor cable to reduce electrical noise in the trigger circuitry. I used 20kV wire and made up boots for the distributor towers from several layers of shrink tubing. I used simple low-profile clips for attaching the wires to the plugs instead of boots because the CM6 plugs are already too tall for the scale of this engine and adding boots to them will increase their heights. After inserting all the plugs I was glad to still be able to feel the compresion bumps when the engine was manually turned over. Having nine cylinders places them pretty close together. - Terry