A/C Evaporator

I bought my A/C kit quite a few years ago from Airflow-Systems. I was shipped a so-called “Australian” evaporator, which is actually a product called a Monster Trunk System, part #685000-VUY from https://www.vintageair.com. There was a collection of metal parts and adapters in the kit, with no obvious way to set up the air flow and no instructions for this evaporator unit. The evaporator contains a 3 speed high volume scroll blower, which is ill suited to pressurizing the overhead console. Several other builders have supplemented this evaporator with an inline blower  which is more capable of pressurizing the overhead console. Yet another technique has been to forget about the overhead, turn the unit around and fit enough ducting to blow air straight into the cabin – see here.

I was already committed to a conventional mounting position, having done the inlet ducts and cutout for the overhead several years ago. What I needed to do was complete the evaporator outlet ducting, including an inline blower to suitably pressurize the overhead console, cabin flood air ducting, and a means to use the rear NACA vent air, via the Aerosport products NACA vent valve. This exercise is complicated by the fact that there isn’t a single right angle anywhere in the system, and it all rapidly turned into a 3D modelling exercise. First though, here’s a description of the evaporator inlet system I put together a few years ago:

The F-1006 bulkhead attachment for air to pressurize the overhead is a difficult area. In order to get enough air volume, a significant cutout is required. This mandates a doubler plate, and there is not much room to fit one. Edge distances, strength, clearances, Aerosport overhead console flange dimensions and screw positions for the rear baggage bulkhead all come into play. I wound up using both a shim plate and a doubler, in order to tie the doubler in with the rivets which secure the top of the F-1028 baggage bulkhead channel. Rivets for the doubler are flush on the rear side (where the manufactured heads are inside the overhead console air space) and flush on the front side where the baggage bulkhead overlaps the F-1006 flanges. I lowered and moved the normal position of the baggage bulkhead top screws, in theory they should have landed right in the middle of the Aerosport overhead lower flanges, in practice they are a little above this point but still easy enough to install.

This whole area is so busy, it is difficult to find a good way to “attach” the required air duct(s) to the bulkhead. It’s also not reasonable to have hard attachment points between the bulkhead and the evaporator/shelf, due to vibration and cracking. The idea of this design is to use a 3D printed bulkhead attachment block to achieve the following:

  • It can be fitted to the F-1006 bulkhead after the top skin is riveted on, sealed with some form of gasket material, and brings the air duct attachment plane clear of the bulkhead.
  • Two long #6 screws act as locating pins for the flange of the main attach duct.
  • A thick/soft gasket or manifold can be used between the rear of this bulkhead attachment point and the main duct, to seal airflow and provide vibration isolation.
  • If (when) the evaporator arrangement changes, a new 3D duct can be printed to mate with the existing bulkhead attachment point.

I use only one hole for airflow into the overhead, the side with more area (the F-1028 is offset from the center). Manifolding air into both sides complicates things and is pointless – what matters for overhead air is pressure, not volume. I wanted electrical connections into the overhead as well, so these are on the right hand side of the bulkhead, and will be sealed off in the overhead.

  • f20a
    f20a
    Cutout for overhead air feed, with shim and doubler plates
  • f20b
    f20b
    Holes for electrical conduits into overhead, with doubler plate
  • f20c
    f20c
    3D printed template for bulkhead attach block, to verify it clears all obstacles.
  • f20d
    f20d
    3D printed ABS bulkhead attach block, bulkhead side.
  • f20e
    f20e
    Bulkhead attach block trial fit. Installation can only happen after top skin riveted on.
  • f20f
    f20f
    A/C evaporator on shelf, with 3D printed inlets.
  • f20g
    f20g
    3D printed inlets are contoured to match the curved front surfaces of the evaporator.
  • f54a
    f54a
    Production inlet parts, made from tough epoxy, to replace the ABS prototypes
  • f54b
    f54b
    Production inlet parts in place
  • f54c
    f54c
    Cutouts done, nutplates in place
  • f54d
    f54d
    I added a stiffener to the front face
  • f54e
    f54e
    Fitting a rubber seal over the original evaporator inlet
  • f54f
    f54f
    Cover plate in place, screwed on. Rubber pads on front of inlets, riveted on and sealant applied
  • f54g
    f54g
    Cutout in evaporator shelf for receiver/dryer
  • f54h
    f54h
    Clamping shroud for receiver/dryer

For the evaporator outlet, I designed a manifold which caters for the following requirements:

  • Fits onto the two irregular shaped outlets on the evaporator, with a simple rubber seal and some screws.
  • Provides an outlet for the inline blower. I used a 4″ blower, because it fits. A 3″ blower would probably also be adequate.
  • Provides a pair of outlets for cabin flood air. These should probably be 2.5 inches each, I used 2 inches because that’s the attachment size I have room for on the front (top) bulkhead.
  • Provides a pair of inlets for the Aerosport NACA vent valve, to feed vent air from outside into the system
  • Provides a place to mount a temperature probe
  • Can be assembled in-place, or if necessary by lowering the rear edge of the evaporator shelf (after removing the support).

The following pictures show what I came up with. I had the prototype fabricated in tough epoxy, and it fitted fine except for an indentation on the top that I made to clear the top stiffener. For some reason my measurements were off, and the indentation missed the stiffener by 20mm. I fixed this up and made some other improvements, and just ordered the final version which should arrive here in another week or so.

The 4″ blower just fits in the required space. I’m mounting it to a metal bracket that will be riveted to the cover plate I made up for the evaporator inlet. Also mounted on this cover plate are three relays (for the scroll fan) and a pwm controller for the inline blower. A wiring harness for this can be seen in the pictures, not properly laced up or secured yet. A high side pressure sensor, and evaporator air outlet temperature sensor, are included. The system controls will be on the overhead, except for the master “A/C on” switch which is on the front panel, pilot’s side. Turning the A/C off (before rolling) will be on the pre takeoff checklist, if necessary it can be re-engaged at some point during climb out. Part of the wiring includes a connector that could be used for a micro-controller that would be capable of climate control, if the rotary switch in the overhead is set to the “auto” position.

The final duct is to go from the outlet of the axial blower to the overhead. I printed some prototypes for this on my own consumer grade 3D printer using a flexible material. Once the shape was correct, I decided to order the production part in SLS Nylon. This will be very strong, but will still have enough flex to effectively detach the evaporator/blower assembly from the airframe. Although I will be assembling all of the final components with the top skin still partly open, everything is designed to be removable and reassemble-able after the skin is in place. It won’t necessarily be pleasant working back there in the hell hole, but it can be done. For assembly, I’m going to take advantage of the skin being off and will cheat as follows:

  • With everything in the tailcone finished, and with the evaporator/shelf removed, cleco the top skin on in its entirety. With Rosie outside on the rivet gun, and me inside, we’ll rivet the holes across the front and towards the rear on each of the three stiffeners.
  • Remove all the remaining clecos, allowing access from each side.
  • Fit the bulkhead adapter and the (flexible) duct from the inline blower outlet to the bulkhead. This can’t be done until after the (above) rivets are set.
  • Install the evaporator, shelf, outlet manifold, inline blower, NACA vent valve etc, using the access from each side to make the job easier this first time.
  • Fit the remaining refrigerant hoses etc. and charge the system. I plan to use an electric motor with a grooved pulley and a long serpentine belt as a means to run the compressor for this step.
  • Check for leaks and proper operation.
  • Cleco the skin back up, climb inside and finish riveting on the skin.
  • evap_out-1
    evap_out-1
    Evaporator outlet manifold
  • evap_out-2
    evap_out-2
    Top view
  • evap_out-3
    evap_out-3
    Bottom view
  • evap_out-4
    evap_out-4
    Right side view
  • evap_out-5
    evap_out-5
    Left side view, small hole is for temperature sensor
  • evap_out-6
    evap_out-6
    Front pump, vent attachments
  • evap_out-7
    evap_out-7
    Rear vent attachments, slot for top stiffener
  • evap_out-8
    evap_out-8
    Evaporator and vent attachment points
  • f56a
    f56a
    Back from the printer
  • f56e
    f56e
    Left side, temperature sensor
  • f56f
    f56f
    Temperature sensor
  • f56g
    f56g
    Aerosport NACA vent valve in place
  • f56h
    f56h
    Left side cabin vent scat tube attachment point
  • f56i
    f56i
    Right side clearances OK
  • f56j
    f56j
    Top view
  • f56k
    f56k
    There's just one problem....
  • f56b
    f56b
    Bulkhead adapter 3D printed in tough epoxy
  • f56c
    f56c
    Bulk attachment point
  • f56d
    f56d
    Bulkhead adapter in place
  • ba2
    ba2
    Bulkhead adapter design
  • ba1
    ba1
    Bulkhead adapter, left side view
  • ba3
    ba3
    Bulkhead adapter design
  • f56l
    f56l
    Printing the bulkhead adapter prototype
  • f56m
    f56m
    Finished bulkhead adaptor prototype
  • f56n
    f56n
    TPU is very flexible
  • f56o
    f56o
    Top view with bulkhead adaptor prototype in place
  • f56p
    f56p
    Checking bulkhead adaptor prototype fit
  • f56q
    f56q
    Side view of bulkhead adaptor prototype

 

 

Inlet plenum progress [2.0 hours]

Construction work has been spotty for the past few months due to some work commitments. Time to catch up on a few posts.

I received the prototype 3D printed inlet plenum, after a very long delay caused by Covid-19, a shipment lost in customs, reprinting a replacement, more shipping delays etc. This part was printed in ASA material by a vendor in Sweden. Quality is good apart from some areas where the wall thickness should have been greater.

The part fitted perfectly around / through the compressor, the air filter shroud mount etc. The inlet ramp was also good, perfectly horizontal and lined up with the opposite side (standard Showplanes fiberglass plenum) within 1mm, which is good enough for me. The front edge of the inlet ramp is too far forward, requiring me to trim too much of the cowling. While it would work, it doesn’t leave as much of the cowling inlet hole as I’d like, to get nutplates and some sort of overlapping seal in there. This is one of the areas where the wall thickness is also a bit low. Clearance on the bottom side to the cowling is good, a bit over 1/8″ at the closest point. The rear scat tube connection for a heat muff is also good.

Overall, close to a hole in one which I’m relieved about given how complex the part is. I can now proceed to finish the front of the baffles and around the governor. I’m going to need to address the wall thickness issue and trim back the front edge, which means re-printing the part. These are simple adjustments but I’m going to also look at whether any better alternatives exist than the ASA material I’ve used.

It’s really hard to see much from the pictures, because the 3D printed part is black, but they show the general idea.

  • f55a
    f55a
    Prototype inlet plenum in place
  • f55b
    f55b
    A/C fitting clearance OK
  • f55c
    f55c
    A/C fitting clearance OK
  • f55d
    f55d
    Prototype inlet plenum in place
  • f55e
    f55e
    Good fit around A/C and into air filter shroud
  • f55f
    f55f
    Good fit around A/C and into air filter shroud
  • f55g
    f55g
    Horizontal, and about 1mm below right hand side plenum - good enough

 

Tunnel heater hose [6.5 hours]

Some time ago I added some brackets to the front of the tunnel, so I could secure the rear heater hose. With Control Approach rudder pedals, the hose needs to be secured in the center of the tunnel, clear of the control arms off to each side of the tunnel. I was going to use a 2″ Adel clamp around the scat tube, based on what another RV-10 builder had done.

After assembling this, I didn’t like it because:

  • It was difficult to install the Adel clamp, while lying on my stomach with the seats removed and reaching in under the panel. The rear heater hose has to be sort-of scrunched against the short front heater hose in order to get it positioned in the middle of the tunnel.
  • The large Adel clamp, held by a single bolt, was not very secure and could have a tendency to rotate over time
  • If anything came undone over time, the compacted rear heater hose would push loose items towards the rear, straight into the rudder pedal arms.
  • I still had to come up with a solution to replace the standard F-1051J Scat tube support, since this support interferes with the internally run rudder cables when the Control Approach rudder pedals are used.

After a few minutes pondering these problems, the solution hit me – design and 3D print a pair of Nylon brackets to retain the scat tube. The brackets then simply slide onto the scat tube from the rear. For the front bracket, I bolted the Nylon piece to the Aluminium angle retainer on the bench, slid it onto the scat tube, lifted the rudder pedal arms, positioned the bracket assembly and screwed it into position. I also drilled a pair of small holes into the Aluminium angle in order to add a safety wire each side, that way if the brackets ever came loose for any reason, the assembly could not fall aft and interfere with the rudder pedal arms.

For the aft bracket, I had already a long time ago drilled and dimpled the holes on the right hand side of the tunnel for the standard F-1051J scat tube bracket. The lower of these two holes is close to the right hand rudder cable. It would have been better to raise this hole by about 1/2″, but that is ancient history. I resolved this by using a low profile (AN364) lock nut and embedding the nut into a hexagonal cutout in the bracket, as shown in the pictures. I used a pair of 0.063″ shims on each side, with the holes countersunk, to complete the assembly.

It all worked great, both brackets can be easily removed and reinstalled, so any future maintenance that requires removing the rear heater hose for better tunnel access will be easy.

PostScript:

A couple of other RV-10 builders have asked me for the models, and one questioned why I elected to use the metal shims on the aft bracket. I used the shims simply because I didn’t think my consumer grade 3D printer could do a good enough job of the countersinks, when printing them in Nylon, vertically. In any case, I added an option to the model to have no shims, which widens the aft bracket to compensate for the missing shims, and adds countersinks to the sides to allow for the #8 dimples in the tunnel walls. I’ve added pictures of this version. The three STL files can be downloaded using the following link:

Download

 

  • front_heater_bracket
    front_heater_bracket
    Front scat tube bracket
  • aft_heater_bracket
    aft_heater_bracket
    Aft scat tube bracket, replaces F-1051J
  • f47a
    f47a
    Front Scat tube bracket assembly
  • f47b
    f47b
    Aft scat tube bracket, replaces F-1051J
  • no_shim1
    no_shim1
    Aft bracket "no shim" version, with included countersinks
  • no_shim2
    no_shim2
    Aft bracket "no shim" version, with included countersinks
  • no_shim3
    no_shim3
    Side slopes to match tunnel

3D printed system brackets [3.0 hours]

Under the front seats, there are four “systems brackets”, F-1084A/B, which have slots for a fuel line, brake line, and electrical wiring. In my case, there are two fuel lines, so one issue is how to deal with the return line. Another issue is the fact that s/s braided teflon lines are different in diameter than the Aluminium tubing lines that the system brackets were designed for. There is apparently enough scope to squash them in with a sliced apart grommet.

None of this sat well with me, and it was a simple matter to design a replacement upper bracket section and 3D print it in Nylon. There are two right-hand and two left-hand brackets, which have snap-in rings for two fuel lines, one brake line, and electrical wiring. Each ring has a slot for anchoring a cable tie, or waxed string tie, if needed. Each part takes about two hours to print, and bolts straight onto the standard lower systems bracket F-1084A. If in the future I need to make a change, I can simply print up new brackets and bolt them in place.

The brackets weigh 7 grams each. This replaces the metal upper section (F-1084B) and three snap bushings, which weigh a total of 9 grams, so there is no weight penalty in this change.

  • f15a
    f15a
    System bracket design
  • f15b
    f15b
    System bracket design
  • f15f
    f15f
    Part being printed
  • f15g
    f15g
    Printing complete
  • f15c
    f15c
    Parts replaced weigh 9 grams
  • f15d
    f15d
    Nylon system bracket weighs 7 grams
  • f15e
    f15e
    Plenty of stretch in Nylon support rings
  • f15h
    f15h
    Four system brackets
  • f15i
    f15i
    Bolts to standard metal parts

3D printing an aircraft

The conventional method for wing wiring in an RV is to run a corrugated conduit from wing root to wing tip. In accordance with Van’s recommendations, I’ve enlarged the tooling hole in each rib, and will run 16mm conduit through this hole, secured in place with RTV. Many find that an extra conduit is required, and run it through rib lightening holes. The issue is how to secure this second conduit. The lightening holes in the wing ribs have an arc shaped recess next to them, so I can’t use the Panduit fittings I used in the empennage. I didn’t like Van’s suggestion of drilling a #30 hole and using a cable tie around the conduit. I looked around for a commercial fitting, but couldn’t find anything suitable.

This problem led me to design and 3d print a suitable fitting, custom made for RV-10 wing ribs. The pictured design holds a 20mm conduit, and has two smaller holes for either air lines (e.g. Angle of attack from pitot) or RG-400 cable (for antennae mounted in the fibreglass wing tip). There is a slot for a cable tie, and the fitting nests into the arc around the lightening hole. It is held in place with two pop rivets, and these are positioned so as to cause no interference with a bucking bar held blind inside the wing, since this is how the bottom skin must be riveted. There is a left and right (mirror image) version of the fitting for the respective -L and -R wing ribs.

Common materials used on consumer grade 3d printers are not suitable for this application. PLA has a low glass transition temperature (Tg) of 66 degrees C, so the fittings would easily deform inside a wing parked in central Australia. ABS has a much higher Tg, and is strong enough, but has poor chemical resistance. An AvGas leak working around inside the wing would cause the part to degrade.

The material I’ve used is Nylon – specifically this product. It is very strong but still flexes under load, has a suitable high Tg of 82 degrees C, and very good chemical resistance. It’s harder to use on a consumer grade printer, but once suitable settings are established, the results are excellent and very repeatable. You wouldn’t use this process for anything structural on an aircraft, but I’m sure there will be plenty of other applications for a 3d printed Nylon part before I’m done with this project.

It takes about an hour to print each part, and I need 30 of them. They weigh 4 grams each.

Postscript: The conduit clips work great, I’ve added some pictures of the assembled wing with conduit in place.

  • wing_clip1
    wing_clip1
    Upper side of conduit clip. One rivet hole is recessed so that an LP4-3 rivet has enough depth to set properly
  • wing_clip2
    wing_clip2
    Rib side. The "bump" fits into the recess around the rib lightening hole. Slot on right hand tab is for an optional ...
  • w19c
    w19c
    The Nylon filament
  • w19a
    w19a
    3d printing in Nylon
  • w19b
    w19b
    Part nearly finished in printer
  • w19d
    w19d
    Part complete, ready to pry off the bed
  • w19e
    w19e
    Position of part on bottom side of W-1011L wing rib
  • w19f
    w19f
    Position of part, showing 20mm conduit, AOA air line, and optional cable tie
  • w38a
    w38a
    Right wing gap fairings complete, wiring conduit fitted
  • w38b
    w38b
    Wiring conduit in right wing
  • w40e
    w40e
    Conduit, pitot, AoA lines in left wing
  • w40f
    w40f
    Checking pitot mast arrangement in left wing