It's safe to say that any flying machine is a collection of parts that are strategically placed and assembled to enable one to fly.
Although this is true, it's the little things that get accomplished over the course of a building project that make the difference in not only getting the plane built, but improves the overall quality and performance of the end product. It also contributes to the philosophy I spoke of last August on my blog: ( a-little-here-a-little-there).
Today I decided to pull out the Corvair heads and have a closer look at them.
You might recall that these heads are 110hp versions from a 1966 car. They are just as I received them from the seller - not filthy but certainly not clean either!
Any engine that is more than 50 years old is bound to be grimy. Air cooled engines like the Corvair have many, many cooling holes factory cast in between the head fins. Being small, they trap everything. Not good as the GM engineers counted on these being clear for optimum cooling. What the GM engineers didn't consider is that casting aluminum or other metals sometimes leaves "flash" where the moulding halves join up during the casting process. Either it was considered to costly to remove the flash or maybe they decided it was good enough. What I have experienced however is the massive range of acceptable "flash" tolerances - some heads have so much the cooling holes are almost closed over. With a small amount of dirt in there, they are effectively closed to cooling air.
Here is a look at the cleaner of the two heads. I've put a lightbulb behind to ease the viewing:
As you can see, the cooling holes are many and this head seems quite good. Some minor flash to clean up but generally good.
The 2nd head, despite being from the same casting lot as the first is terrible! I could only find one hole that was clear and even then it was almost closed over from flash:
Took an hour with a combination of small files and old steak-knives, but I managed to clean out the majority of the gunk from the cooling holes.
Both heads really do need a good pressure wash before I'll be able to remove all the flashing, but this was a good start. Removing the flashing is very important and goes a long way to improving cooling of the heads.
It's these little things that take time, but make a world of difference and counts towards the goal. Along those same lines, my father Jim and I have been working on an important submission regarding my mentor Barry. I don't want to say much yet but it's a little thing that also counts for something. Stay tuned for more on this shortly.
Yup, I said it..... no, I can't speak german... thanks Google translate.....ha ha!
The Corvair authority William Wynne talks extensively on his blog about different carb applications in a Corvair conversion and the importance of keeping things simple.
The dual (and sometimes quad) factory Rochester carb setup on a Corvair car engine not only complicates matters (syncing throttles arms, etc) they were never designed for altitude compensation and mixture settings required in an aviation application. The converted Corvair engine falls into the same horsepower range roughly equivalent to typical medium Continentals and Lycomings, approximately 100 to 120HP. This requires a fuel delivery system capable of delivering an air to fuel ratio capable of supporting this demand.
Fuel injection? I believe the advantages (no carb icing, small increases in HP) are FAR outweighed by the complex system components (injectors, return fuel lines, pumps, electronics, sensors, etc). Keep it simple.
The MA3-SPA carb as found on the O-200 continental and O-235 Lycoming is the definition of simple. They haven't changed much since the 1940's and Marvel Schebler continues to make new ones today - in other words it works, simple. Overhauled to new specs it's the perfect carb for my conversion.
Finding one that is both inexpensive to obtain and overhaul becomes a problem due to this popularity. A good core for rebuild can be found in the three to four hundred dollar range then count on six to seven hundred dollars to overhaul it. Expensive, but not an area I want to save money on - engine reliability is important in flying! The recommended overhaul shop (D&G Supply in Michigan) also will convert specific O-300 carb models to the Corvair specifications.
Armed with this knowledge, I've been searching online for a suitable core.
A couple of weeks ago while surfing E-Bay, I came across a listing for an O-300 Marvel Schebler carburetor that would be suitable for my engine. It's clearly an older one, but again the model number matches the acceptable models for conversion and the pictures showed well.
Like anything on E-Bay, Kijiji or Craigslist it's a buyer beware mentality. One has to consider the odds and what it's going to cost to ship. In my case, the core I was interested in had no reserve pricing but the shipping costs weren't cheap - it was in Germany! This compounds the pricing with the Euro being somewhat strong against the Canadian dollar. Worth a shot.
With this in mind, I did the responsible thing and figured out my maximum bid would be about 100 Euros. I watched the days count down and was pleased to see my bid of 40 euros was enough to win! With shipping and currency conversion the total costs came to $112 Canadian. Not bad and certainly better than what I expected to pay for a core.
"Mein vergasser ist angekommen" (which means "my carb has arrived") on Friday and I picked it up at the post office yesterday. My first look had me really worried as the box had a crushed corner and was split open at the top:
There was a sticker on the box from Canada Post stating the box was damaged by the forwarding shipper. Not good. One of the fears shipping any item overseas or otherwise is theft. Hope there isn't just a bag of sand in here!
Opening the box, I smiled a bit finding a note from the seller:
Opening the box further, it came apparent that the shipper used a lot of bubble wrap to protect the carb on it's journey, but more importantly there is a carb inside the wrap!
It's definitely old, but everything seems intact and the throttle/mixture arms move freely. The accelerator pump seems seized but that's typical of something that has been sitting on a shelf for a long time and that will be repaired as part of the overhaul.
The data plate is intact and shows this is a model 10-4895 MS carb, typically used on O-300 engines. This is a good carb for overhaul and conversion to the required specs for my Corvair!
Glad I found this. It will be sent for overhaul this fall.
Back to the shop soon.
So, I got my block/case back from the machine shop. It was worth every penny to have a professional with a CNC milling machine do the work of drilling out the two broken studs. His work was incredible and he went as far as to countersink the holes slightly for the TimeSert barrel inserts. Nice, clean and straight holes, important details:
With that done, it was time today to tackle installing the TimeSert barrel inserts that will make up the replacement threads for the head studs on these two holes. I've spent more hours than I should have pondering this critcal step, but it wasn't nearly as difficult as I allowed my imagination to believe it could be.
TimeSerts are an elegant solution for replacing damaged in-hole threads in a variety of materials. They are in my opinion much better than Helicoil wound wire inserts. I ordered the TimeSert install kit from Clark's Corvairs and the recommended length (0.75 inch) TimeSerts from a local industrial supply shop. The kit contains all the tools you need to install these:
First step, drill out the hole to the correct size. The importance of having this hole straight can not be understated. Although drilling aluminum is easy, best to use a drill press:
The brand new bit that came with the kit was very sharp and made short clean work of the holes.
Next up, the shoulder countersink bit. The countersink the machinist put in was quite deep enough. The bit has a cutter which creates a countersink shoulder for the top end of the TimeSert allowing it to sit flush on the surface. Again, the drill press is the only smart way to do this:
Careful application of preasure on the very sharp cutter results in a nicely formed shoulder.
Next up, threading this aluminum hole with the tap provided in the kit. This tap (also brand new) has four cutting flutes and a flat nose to ensure the hole is completely threaded to the bottom. This is delicate work that is only done by hand, so it was important to make it perpendicular to the case, ensuring straight threads in the hole. I used lots of 3-in-1 oil to ease the tap through and keep the threads clean from debris. The secret is to cut 1/2 a turn, back out 1/2 a turn and repeat:
Once both holes are tapped, it's important to ensure they are cleaned out of any cutting debris. A blow gun and compressor is perfect.
At this point everything looks good. Next up is the insert mandrel.
What makes TimeSerts so effective, is their engineering. The bottom couple of rows of the insert are formed in a way that allows the insert mandrel to cold-roll the threads, pushing them outwards and into the surrounding material, locking the insert in place:
First, a couple drops of oil on the mandrel to ease the insert forming the new threads, then thread the insert partly onto the mandrel:
Once placed in the newly cut holes, the insert threads in easy, up to the shoulder stop. Continue moving the mandrel forward (in) until it bottoms out:
Back the mandrel out and the TimeSert stays in place, now permanently attached to the aluminum walls of the hole. Looks perfect!
Was it really that easy? YES!
After cutting the matching threads on the two upper (long) studs that will be used with these inserts, I test fitted one. A good clean fit that will be made real strong with LocTite 620 as per the conversion manual.
I can't explain how relieved I am to be past this part of the head stud saga. Next steps will be fastening all the studs in permanently with LocTite 620. Before that, I need to clean the block completely as there are still areas that can use some detail attention. The suggestions are to either hot tank the block at a tranmission shop, use varsol with stiff brushes and 3M pads or maybe media blast it. Ron has a sandblasting machine, maybe that would be easiest. More food for thought.
The other task I completed was the removal of the oil pick-up assembly. It is press fit into the aluminum:
A little gentle tapping from the oil passage end of the block witha pirce of dowel, and it came right out:
Overall, a very productive 3 hours. I'm tagging this blog entry as a milestone because it has been bugging me for months to get past it. Barring any further surprises, this will be my airplane engine.
Now back to making other airplane parts :)
Looking back I can't believe it's been almost a month and half since I posted to the blog. I've been busy waiting on some stuff I ordered to take the next steps on the motor rebuild and some travel to visit family in Ann Arbor Michigan took up a bunch of time as well. Well worth it though, we needed a quick family get-away to recharge.
Since my last post, I took my engine block into the machine shop to fix the snapped stud issue. They have a highly accurate CNC milling machine which will make short work of the snapped stud. The process will remove the remnants of the broken stud but this will also mean sacraficing the hole threads. I've now got the TimeSert install kit that I ordered which will repair the damage and create a new set of threads to insert the studs into.
Until I get the block back from the machinist early next week, I got some of the prep work needed for the studs done. The original short (lower) head studs from my core engines are in decent shape, but typically very dirty with some light surface rust. The push rod tubes are similar. Here is the before pic:
The perfect tool to clean these is the bench grinder. This one is a beautiful old school one. I prefer old tools that are made to last:
The wire wheel makes short work of cleaning of the decades of old grime and rust and is excellent for cleaning up the stud threads:
The push rod tubes from the core are really dirty. Under the grime, the tube was manufactured with two coatings on top of the bare metal, as shown in this picture from Google:
I had a go with the wire wheel on one of my tubes and this was the result:
I'm happy with how they cleaned up, particularly around the o-ring area. However, after seeing the picture from Google that I found for tonight's blog entry, I'm not sure if I've just removed the grime, or removed the zinc coating as well. Removing the zinc coating and getting down to the bare steel is what I want to do as this will allow me to paint (or maybe powdercoat) them white as described in the conversion manual. I'll have another look next time I'm in the shop. These may require a bit more work. The one in the picture I copied was sandlasted, perhaps that's what I'll end up doing.
Overall, the first twelve studs all came out really nice and clean - they should paint up real nice. There are some minor tool marks on each. I have a bunch more in the inventory, so I'll clean those up too and choose the best ones for the build:
The other task I've been pondering is cutting new threads on the end of those studs being inserted into the TimeSert holes of the block.
The studs on Corvair engines are made from a very high tensile steel alloy. The original threads at the block end are a proprietary GM thread called 38-16 NC5. These will not fit the TimeSert which are the more common 3/8-16 NC. So, for those holes that I'm installing TimeSerts, I'll have to use a die cutter and rework the threads to be 3/8-16 NC. The head end doesn't need to be altered.
I had a bit of time today, so I took one of the old long (upper) studs that are being replaced with new ones due to corrosion and experimented cutting new threads on it. Best case, I see how easy or difficult it is, worst case I ruin an old stud that I won't be using anyhow.
I clamped the stud tightly in the vise. When I do the good ones, I'll have to remember to put something in the jaws of the vise to prevent clamping damage marks. Here is a picture of the tools I used. I couldn't find where ron keeps his cutting oil (if there even is any), so I substituted a little 3-in-1.
There ins't a huge difference in the GM thread and the 3/8-16 I need to use with the TimeSerts. Carefully starting the die on the threads and using a fair amount of oil, I managed to cut or reshape the threads about under half way down. This involved the time proven method of turning the die down a little bit at a time, and backing off numerous times but it went marvelously well. Here's a close up:
When I got home, I test fitted this stud in a TimeSert and it threaded in real nice. I bit of LocTite 620 should make the repair as good or better than the factory fit. I've been worried about this process for a long time, but I think with a little attention and time, it's going to work out fine.
Next up, prepping the block for stud install.
The Google search bots are really going to love my posts now!
Remember my fellow Corvair engine builder Jeff Moores of Newfoundland (see previous post "time-to-get-back-at-it")? While at the Zenair Open House we talked over lunch about the struggles I had been having with head studs and Jeff reassured me that my issues were common issues in both his previous builds. He offered to send me some extra head studs that he had lying around his shop to replace the bad ones from my core. They arrived via mail on Tuesday and they are brand new! All for the price of shipping via snail-mail.
The more I continue pursuing a Corvair as my choice of motor, the more I'm starting to realize the value of getting to know other Corvair builders, both for their experience and generosity. This is the kind of group I want to associate with, not some faceless foreign owned engine maker that just wants my money and couldn't care less about my mission to learn. Thanks Jeff!
Next steps, dealing with the 3 stud holes that need to be fixed (see "progress-sort-of") . I've decided on using TimeSerts which are a threaded barrel insert repair that is accepted in the conversion manual. Definitely more expensive than Helicoils (another possible repair method) but I believe worth the piece of mind. Corvair automotve parts warehouse Clark's Corvairs rents the TimeSert installation tool kit and also sells inserts that are the proper length and a blind nut tool for proper torquing of the head studs. I think I'll order those now and get the repairs done soon in preparation for some case machining work I'm planning.
I haven't posted anything to the blog because I've really not made any progress in the past couple of weeks, other than research stuff. Summertime is difficult.... hot and muggy and plenty of other family plans and activities to occupy my time.
Had a few minutes today though, so I thought I'd get out the core crankshaft I have safely stored in inventory and have a better look at it. One of the first items I'll be sending away for prep work, but it needs to be measured to see if it still meets factory specs.
This is the crank I inherited from the inventory I purchased in February. The previous owner had already completed some of the work to make it airworthy and it's in great shape:
The recommended prep work includes magnaflux testing for internal damage, straitening if required, heat treating for improved strength (ion nitriding) and grinding the piston rod journals to improve the fillet radiuses. Larger journal radii help prevent stress riser cranks from forming, which has led to broken crankshafts in the past. The process is quite common in certified aircraft engine cranks, so it's worth doing here.
One of the other processes that normally gets completed is to drill and tapping the centre of the end of the crank for a propeller hub safety shaft. Because the previous owner was working from the official plans, he had already completed this process with his machinist:
I opened my Corvair shop manual, and found the engine spec reference page. It contains all the measurements for the major engine components:
Using my digital caliper, I tried to measure as best I could the crankshaft main journals and connecting rod journals to see how close they are to factory. I really should be using a micrometer for this, but my calipers should at least let me know it's in the ballpark.
The manual says the connecting rod journals (they call them crankpin journals) calls for a diameter from 1.7999 inches to 1.800 inches. To make it easier to measure, I set and friction-lock my caliper to 1.7990 before placing it on the journal. This is slightly undersize, but as close to 1.7999 I can get. This way if the caliper slides over the journal without resistance, I know that it is below tolerance and no good for grinding. Anything larger than that leaves that much more room for the machining (good):
With my lovely assistant Brenda taking an action shot with the camera, I carefully measure each of the bearings:
It's real hard to get a good picture, but the caliper won't span the largest diameter of the journal, so we can deduce that they and the main journals (using the same measuring technique) are clearly above the minimum spec and can be used for conversion. I know it isn't a perfect measurement and my caliper may not be as accurate as a micrometer, but I think I'm in the ballpark at least.
I'm going to see if I can borrow a good micrometer from someone, or maybe buy one for myself to confirm this. Once I'm sure, off to Florida the crank will go for prep.
Next Wednesday I'm having my first shop lesson with Ron, another Zenair builder. He's rebuilding a Zenair 701 and has offered to let me help. What a great opportunity to learn and prep for my build.
More to come.....
Took an hour today to start cleaning up the push rod tubes that I removed from each of the cores:
First step was removing the old dry and cracked o-rings. GM used simple rubber ones which in the conversion process are replaced with Viton rings. Viton doesn't dry out with heat and remain supple over the course of the engine's life. Some were already cracked and missing pieces:
I rummaged around in my tool box and found the perfect thing to use. Don't ask me if I know how sharp the point is:
Some came off complete, but most came off in brittle little pieces:
Next step will be to clean them up. First a bath in Simple Green to get rid of the grime and grease, followed by a polish. The tubes are made of light steel with a zinc coating. GM left them bare, but standard practice for conversions is to paint and bake them white with high heat enamel to enhance cooling (or at least reduce heat transfer to the tubes from the cylinders). I've read online it's best remove the zinc coating and this can be done using an aggressive Scotchbrite pad or bench grinder wheel.
Well, my buddy Guy had some progress in getting the last three broken studs out on the 110hp block.
The first one came out fairly easy once centre drilled. Looks like the threads are still intact and should clean out nicely:
Number two, not so much. The centre of the stud hole doesn't always line up exactly with the threads, so he stopped working this one with the end mill until we can decide if we want to re-tap the hole afterwwards:
I tried using a large EasyOut bit but the last of this stud won't come out and I'm wary of damaging the remaining threads. I've filled up the hole with some home-brew penetrating oil and will let it sit for a couple of days to see if it will loosen up before trying again. I think I might be able to use a tap and just clean it out, but I'll wait for now on this one. This is likely the worst of the bunch.
Number 3 was the most interesting result. Using an EasyOut, Guy backed the stud out carefully, but unforuntately it brought most of the aluminum threads with it. It was stuck in there real good:
Now, the beauty thing of all this is that all 3 holes are salvageable. There is enough room left between each of the damaged holes and the cylinder bores to insert either a Heli-Coil or TimeSert thread repair. They each have their strengths and pricing differences and from what I've read they each work well (click the links to check them out). I kinda like the TimeSert method better, just seems more permanent. This will mean cutting new threads on the end of the replacement studs with a sharp die to match the repaired holes, but that is common practice and an acceptable repair for a conversion.
If I can't get the number 2 one out cleanly, it will be easily removed during the drill/tap/insert process of the thread repair.
I'm stoked this is going to work and I can use this block for my engine <grin>.
On another note, I've added a running total time log on the lower right of my blog page:
One of the things builders like to keep track of is the amount of time spent on parts of the project. I've broken things down into basic groups and will try an remember to update it regularly. I considered purchasing a commercially available program like KitLog, but I prefer this blog format. The 25 hours showing under Engine is just a rough guess and doesn't include non hands on time like time I've spent online and in person looking for a core. I might consider breaking the chart down further, but I'm happy with it for now.
Next step, research where to purchase TimeSerts.
Got a call from my buddy Guy (correct pronunciation is Gee, which is french Canadian) . Despite his best efforts to remove the broken studs he is struggling a bit. He tried welding a nut to them but they just snapped off further down and now they are sitting close to flush to the block. This leaves no option but to drill them out using a milling machine and end mill bit.
Fussy, temperamental work with a fairly high risk of wrecking the threads if not careful.
After finding limited success using the weld method, he tried centre drilling the stud in preparation to back them out with an EasyOut bit. This proved to be very difficult because the studs are a hard material to drill, but he did manage to centre drill one of the three. The other two he's going to use the milling machine as it should be easier.
The next concern will be how to clean out the remaining debris from the threads that gets left behind. The conversion manual is very specific that the lower end of the studs is a special thread called 3/8"-NC5. So at this point I believe I'll need a 3/8"-NC5 tap to clean out the threads. Even an experienced machinist like Guy had never heard of this particular thread (he checked with his suppliers too) and suggested it will be expensive to obtain due to it's rarity.
While my buddy worked on end milling the holes, I decided do do some research.online. Although I wasn't able to find a tap or die that matached this unique thread, I did come across an online archive of GM production drawings that show the machining dimensions of a Corvair engine! I really love the internet!
This is where it gets a bit more confusing. According to the dimensional drawing showing the machining instructions of the casting, the stud holes are supposed to be tapped to a dimension 3/8"-16 UNC:
But how can that be? The hole and stud should be the same thread as the stud..... hmmm.
I sent an e-mail to the internet Corvair conversion forum seeking some guidance.
Not long after sending the e-mail, I got a telephone call direct from William Wynne (the Conversion Manual author and recognized Corvair guru).
We had an almost hour long conversation about the conversion process, my overall plans and this particular issue regarding the studs among other things. He's very supportive of new builders like me that want to learn and his overall philosophy about home-building and being in the arena speaks to me.
He is an amazing person to speak with and very quickly confirmed that the GM drawings are correct, the stud holes are in fact 3/8-16 UNC. The reason the studs are slightly different is GM engineers wanted an interference (extremely tight) fit to ensure the studs would remain in place. Using a common 3/8-16 UNC tap would be appropriate to chase the debris from the holes.
Guy happens to have that tap (it's common) and I called him afterward to confirm what we know now to be correct. He's going to finish cleaning things up. In the meantime, I'll bring him the other half of the block and have him clean it up too. On the advice of William I'll also be contacting Dan Wesseman of FlywithSPA.com, William's recommended supplier for info on obtaining 12 new (to me) OEM matching long studs as the ones I have are too corroded to re-use.
To make things even better, Brenda tells me a fellow Corvair conversion builder called for me while I was at work and invited me to visit his shop near Barrie. He is building a Zenair 650 with a Corvair engine. It is almost complete at this point and offered to help answer any questions I might have along the way. I'll make contact with him tonight when I get home and maybe arrange a time to visit this weekend when I'm in the area for a family function.
Despite soaking those stubborn snapped studs in "homebrew", there is too little of them left sticking above the surface of the block to grab onto with a tool and back them out. For all I know, they'd come out now, but I got nothing that can grab onto them strong enough to turn.
I met up with my retired 911 buddy who is an accomplished machinist/welder/hobbyist to have a look at them.
He's taken my block back to his place for a couple of days to see if he can get them out. He figures he's got a couple of options to remove them and preserve the threads. I've told him to take his time and get to them when he can and he promised to be careful. Now that he sees my dilemma in person, he understands. He's the type of guy who likes a challenge and I have little doubt he'll be successful.
Time until takeoff
Husband, father and 911 dispatcher. Long time pilot with a licence that burns a hole in my pocket where my student loan money used to be. First time aircraft builder. Looking to fly my own airplane.