Archive for May, 2013

Heat muff and hoses

Monday, May 27th, 2013

Since air-cooled airplane engines don't have a ready supply of hot water to run a heater core, the only way they can provide warm air in the wintertime is by taking it from the outside of the hot exhaust pipes. This involves the use of a cylindrical heat exchanger, also called a heat muff, placed around one of the pipes and connected to inlet and outlet tubes, which are usually SCAT hose.

One of the little gotchas about the IO-360-M1B engine I have (parallel valve, horizontal induction) is that the only good place to position a heat muff is on the #1 cylinder, which happens to be the one farthest from the firewall and thus requires a long and convoluted hose run. Such is life. The other fun thing is that the heat muff that came with the firewall forward kit didn't fit at all, so I had Rick Robbins weld me up a new custom one. Here's the old one in the foreground, with the new one behind. The new heat muff has the inlet pointing the other direction; the inlet is also downsized from 2" to 1.5" in an effort to reduce the airflow velocity and hopefully thus allow greater transfer of heat.

Since there were no mounting holes in the aluminum flange I bought, I used my mill to make some. I swear, I get more use out of my rotary table than I ever thought possible. It's perfect for little jobs like this.

I cut a hole in the #1 cylinder baffle, and match drilled mounting holes for the duct flange:

Another view from below… locating this correctly is trickier than it seems, due to various clearance issues that are hard to photograph but become apparent when you actually have the parts in front of you.

Test fitting the heat muff and trimming the hoses to length. Inlet air comes from the front baffle, goes down into the bottom of the heat muff, flows upward past the hot #1 exhaust pipe, and exits into a long awkward hose run that goes around the #3 pipe and up to the firewall, where the heat valve is located.

I packed the inside of the heat muff with stainless steel kitchen scrubbing pads, with the intent of further slowing down the incoming air and giving it more surface area from which to transfer heat. I heard about this trick from VAF, and it seems like it could be a good idea for increasing the heater output. Shouldn't hurt, anyway.

The amount of clearance between the hose and the #3 exhaust pipe is adequate, if not ample:

Except, oops, there's nothing to stop the heat muff from rotating around the exhaust pipe, which could allow it to twist as shown here and put the hose in contact with the hot exhaust pipe. Not to mention, allowing the heat muff to spin back and forth will probably eventually saw a hole in the exhaust pipe. Not so good.

I solved this problem by modifying the heat muff design a bit. Each end cap is assembled from two stainless steel half-moon shapes, and then the opposite end caps are tied together with tension rods. I wanted to allow the two parts of each end cap to cinch down tighter on the pipe when the end clamps are tightened, so I used the mill to elongate the holes into slots.

Now the heat muff fits much tighter around the pipe when the band clamps are tightened. It's still not completely fixed in place, but it now takes much more force to rotate it. Seems good enough.

I used a series of adel clamps to secure the 2" SCAT hose where it passes through a jungle gym of engine mount tubes:

Another view from below. The clamps keep the hose from rubbing on the engine mount, and also help aim the hose through its arc around the #3 cylinder.

I installed a couple of radiant heat shields on the #3 pipe where it's closest to the SCAT hose. Another thing that I figure can't hurt:

I laid a bead of RTV along the length of the SCAT hose, which is supposedly an old mechanic's trick intended to keep the hose from totally coming apart if the string starts to unravel.

More RTV:

To keep bugs and other critters out of the heat system, I installed a piece of window screen between the baffle and the duct flange:

Close up of the inlet screen:

The included instructions didn't explicitly say to safety-wire the band clamps that keep the heat muff together, but it seemed sensible and was easy to do:

I squirted a bead of RTV into the gap between the inlet hose and the oil sump, since the clearance is pretty tight there:

And some more RTV between the hot air hose and the engine mount, just in case the clamps want to shift. I think I've gone through three tubes of RTV just in the last few weeks.

With the lower cowl installed, there's about an inch and a half of clearance between the SCAT hose and the cowl. That should be just enough. I don't know how the stock heat muff would have ever fit, since it has the hot air outlet at the bottom.

I think the heat system is complete now. Just in time for summer to get underway!

Installed and sealed baffles

Monday, May 27th, 2013

One of the chores I'd put off earlier, since I knew I was going to have to remove the baffles to tear down the engine, was the messy job of sealing all the small gaps between the baffles and engine with RTV. Well, now that the engine is all put together again, it's time to bolt on the baffles and get messy.

The #3 cylinder baffle (aft starboard corner) was already installed from a previous job:

I discovered a rather disconcerting problem with the baffle on the #4 cylinder (aft port corner). It seems that I ended up with a gap between the baffle and cylinder barrel at the junction with the cylinder head, due to a deadly combination of carelessness, cluelessness, and the prepunched holes being drilled in a less than ideal spot. You can see the gap as a bright spot in the left-center of this photo, revealed by a light placed behind the baffle. This clearly won't do, since lots of precious cooling air will end up spilling through that gap instead of going through the cylinder fins and doing useful work.

The complete solution to this problem would be to remake the inboard portion of the #4 cylinder baffle, which I still may do someday if it ends up being a problem. But in the meantime, my partial fix was to use some scrap aluminum angle to fabricate a little sealing plate, to cover over the worst part of the gap, and seal up the rest with RTV as best I could. I don't know if this will entirely work, but it was relatively easy to do (or at least as easy as working in the confined spaces on and around the engine ever is). In this photo you're looking forward and to the right; the bolt captures the plate, the spacer holds it in place against the baffle, and the glue does the rest.

The prop governor line passes up through the front starboard baffle, through this grommet which was a royal pain to install because the stackup of materials is so thick here. Next time I might think about fabricating some kind of seal plate with a smaller grommet in it, which could more easily installed with a couple strategically placed screws.

At some point during this process we had an unexpected snowstorm (in May!) that caused Mary to haul all of her garden containers into the garage. I thought this was a funny picture:

When RTV has to go into a particular location without turning into a total disaster, I prefer to use a syringe to direct the flow. The 6 mL size is just right; I buy them at the farm store up the road.

RTV applied between the #1 (front right) cylinder baffle and the engine case:

Between the #2 (front left) cylinder baffle and the engine case:

A small spot at the outboard corner of the #2 baffle where it attaches to the cylinder head:

Filling gaps at the inboard corner of the #3 cylinder baffle where it meets the crankcase, below and behind the oil pressure regulator: (reflections make it look messier than it is)

At the outboard corner of the #3 baffle where it mates to the cylinder head: (reflections again)

Filling a gap where the shape of the baffle doesn't completely match the junction of the #4 cylinder barrel and the crankcase:

I installed the little splice plates that join the forward and aft baffle halves at the slip joints:

And, installed the baffle tension rods:

The baffles are now installed on the engine for good! They should not have to come off again unless I discover some kind of unexpected cooling problem, or if I have to tear down the engine again. Please, let's not do that again already, I can't afford to have that much fun.

EGT/CHT probes

Monday, May 27th, 2013

This photo shows the four exhaust riser pipes, each with a carefully-located and potentially very expensive hole drilled in it. Each one is 3.5" inches from the mounting flange, and oriented in a certain way that I determined after some tedious measurement and trial fitting:

Into each hole goes an exhaust gas temperature (EGT) probe, which is a clamp-on thermocouple that looks like this. I shortened the band clamps on each one to remove extra material I didn't need – these must be sized to fit some kind of monster exhaust pipes. (original version on the right, shortened version on the left)

The EGT probes clamp on to the exhaust pipes like so. I was skeptical that this would provide an adequately tight seal, but they are quite a tight fit into the reamed holes, and I don't think any exhaust blow-by will get past them.

Close-up detail of the safety wire on the clamps. They want you to tighten these to a surprisingly high torque – well beyond what you can do with just a screwdriver – and you definitely don't want them loosening up.

Exhaust reinstalled on engine, complete with temperature probes:

Meanwhile, another set of thermocouples is threaded into the underside of each cylinder, to allow the pilot to monitor cylinder head temperature (CHT). Here you can make out the CHT probes for cylinders 1 and 3:

Here are the CHT probes on the opposite side, for cylinders 2 and 4. (Pay no attention to the state of the mixture bellcrank in this photo, I finished re-hooking it up after I took this)

EGT and CHT probes installed on the starboard side, but not wired yet:

From this angle you can see why measuring the locations for the EGT probes was tedious… you have to make sure the sticking-out portion of each probe is tucked in enough that it doesn't get close to the cowl, but is pointed outboard enough to ensure that it doesn't interfere with anything else on the engine:

Probe wires in disarray on the port side of the engine. In time, I will bundle up all the thermocouple wires and extend them aft to the firewall. But before I can do that, I need to install the magneto and associated plug wiring. One thing at a time.

#3 cylinder cooling modification

Monday, May 27th, 2013

Yes, I am still working on the airplane! My excuse for not posting any updates for a while is that I've been ridiculously busy at work, and also I've been jumping back and forth between separate but overlapping firewall forward projects that haven't really lent themselves to turning into a single writeup. So beware that the photos accompanying the next few posts may show multiple separate things happening at once. Don't panic.

Here's the back of the engine, viewed from the starboard side. The cylinder closest to the camera is #3.

Here's a closeup of the back of cylinder #3, which is the intake side. The issue on the RV's is that the cooling fins here are really shallow, and in fact are partially blocked in places by the design of the cylinder casting. The RV baffles sit hard up against the aft face of the cylinder, which has the effect of choking off a lot of the air that is unable to pass from the top of the cylinder to the bottom through the blocked fin passages.

Here's another view with different lighting, in which you can see the problem. The fin area at the center of the photo is very shallow, and the portion of the fins in the lower right (near the intake valve) is blocked entirely:

The traditional approach to ensuring adequate airflow through the cooling fins is to add a washer between the baffle and the cylinder, at the screw hole shown on the right side of the previous photo. This probably works, but opens up a gap along the entire back side of the cylinder, which wastes cooling air. A better approach is to form a duct in the baffle to let air "jump over" the blocked fin passages. Since my baffles were already finished and fitted, I chose to address this by riveting some strips of 0.090" scrap aluminum to the inside face of the baffle. This isn't as good as a formed duct, but it was easier to accomplish:

Alternate view showing the spacer strips. The inner portion over the cylinder barrel is blocked off, since the fins there are plenty deep, and the outer portion is blocked off since the fins are either adequately deep or nonexistent.

Here's what the channel looks like when the baffles are installed on the engine. Air can go down through the channel formed by the riser strips, but is forced back through the lower cylinder fins by the circular wrap that goes underneath the cylinder. As a result, the air is able to do some useful work cooling the engine, instead of being blocked. To the left and right, the air goes down between the cylinder fins as per normal.

Viewed from underneath, you can see the vertical spacer strip that's necessary to keep the air from squirting outboard instead of staying between the cylinder fins and continuing downward. Although it looks like some of the lower outboard fins are "orphaned", i.e. cut off from cooling air, the design of the cylinder head and stock baffles would have stifled these vestigial fins anyway, so I don't think any cooling capacity is lost by doing it this way.

On to the final baffle installation. This was an easy mod to do, once I understood the geometry of the airflow and the nature of the problem to be solved. If it helps prevent cylinder #3 from running too hot, great – otherwise, it should be no big deal to remove these spacers later down the road if I need to address this area again.