Engine Cooling Ideas

This blog covers the background behind air cooling design

Overview
With the aircraft advancing and beign at home for a few weeks over Christmas it was decided to firm up the cooling of the cowl.

Why?

Simple, I would like to close up the hole under the nose gear but this is currently the primary means of cooling and engine air other components will be required.

The radiator is the primary concern as it size is fixed and options are few. Let get this straight it has worked in the factory aircraft for 300 plus hours and there are others on the way but one effect of closing a cowl is the internal temperatures climb and air has to have a routed in/out.

The first blog simply says you need ducts so that easy just build an internal cowl. The original design showed a connection to the fake exhaust - nice but not practical as to look correct they have to reside on the fuselage and the connection would be a bear.

What was needed was a means of extracting the air while maintaining the cowls look and this restricted the ideas to one - louvers. After a lot of work I have settled on a set of automotive louvers 5'' x 6.5 '' overall for a WRX.

Proposed duct - 550 mm overall

The inlet in light orange will be installed as the airflow is underrated in climb remembering that engine and oil cooling have to be handled by the radiator and the calcs show the radiator and water flow underrated this mode. The heat exchanger for the oil will be incorporated into the engine shroud for a number of reasons but one is to allow the exchanger to have an air blast to help offset load on the radiator.

The front duct will decelerate the incoming air raising the dynamic pressure with the duct on the reverse side accelerating the air reversing the process and then raising the pressure at the exit again. 

A wicker is just an additional trip at the front vent
The airflow behind the vent would be interesting

The other reason is that this mimics the pressure under the bonnet allowing air to exit into the moving stream under the cowl. The vents shown have a adjustable lips that extends up into the air flow effectively tripping the air stream creating a low pressure zone allowing the air in the duct to efficiently enter the moving stream of air. One [1] vent will be installed on the starboard side matching the SCATT hose.

RV12 Radiator / oil cooler duct

Another vent will be placed on the port side to vent 12 kw of heat created by the cylinders and radiant heat from the exhausts, radiator and other items. 

Note: This will only be effective with a sealed cowl


The fins are cooled by a downdraft pendulum used on the RV12, this will be covered in a later blog but in the heat exchanger plates located on its underside will be installed into the shroud inlet using incoming air to cool the cylinders taking some load off the radiator as the 64 mm diameter inlet has an excess of air available.

If door are installed they will have two sets of louvers fitted to allow the radiated heat to be ejected on the underbelly [see note above]

The final item examined was the supercharger air supply as this is also part of this equation. A 75 mm od supply was used for a simple reason, its what is available for the aftermarket automotive market. The inlet and filter are both carbon fibre and very light, cost effective with all connection being achieved using 75 / 64 mm od SCATT hose.

Comment

The rough estimate of the workload outlined below shows that there is a potential load at full power of 48 kw on the radiator with the water pump providing 50 Kw of water with 65 Kw of air available in level flight. If correct it means the system has to be very efficient in climb and when full power is beign used as every cubic cm of air must be harnessed along with the cooling water and there is a need to verify the capacity of the radiator in the real world carefully!. 

Wishing I had paid a lot more attention when in sitting in the classes - always wise after the event so below is some rusty thermodynamics.

The initials calcs were based on a article for the Europa by Jans
Apologize for any error in advance

Maximum power output Rotax Supercharged - 100 kW

Consumption at maximum power 40 l/h, i.e. 11.1 ml/s.

Gasoline represents 35 kJ of energy per ml.

Power consumption is therefore 11.1 x 35 = 388 kW

Efficiency is 100 / 388 = 26%.

Maximum cruise power output = 79 kW

Consumption at maximum cruise power is 29 l/h, i.e. 8.0 ml/s.

Power consumption is 280 kW

Efficiency is 28%.

75% cruise power output  55.1 kW

Consumption at 75% cruise power is 20.4 l/h, i.e. 5.67 ml/s.

Power consumption is 198 kW

Efficiency is 27.8%.

Total heat production to be removed estimated :

388 – (85 * 1.13) – (12.5 * 1.13) = 278 kW.

Assumptions

Rotax 914 heat removal prescription for maximum power operation (86 kW)

has to reject 45 kW as follows:

30 kW through cooling radiator

9 kW through oil radiator

6 kW from cylinder barrel fins

Supercharger factor = 912S/914 = 105/86 = 1.22

Under the cowling.

As the engine and exhaust system heat up they radiate more heat.

Stefan-Boltzmann says: about 5 / 10^11 x T^4 kW/m^2, T in Kelvin.

Table - T in Kelvin ('F) and corresponding radiation flux in kW/m^2 - :

700 (800'F) 12.0

800 (980'F) 20.5 - supercharger runs about 200 degrees cooler than turbo

900 (1160'F) 32.8

1000 (1340'F) 50.0 - turbo calcs

1100 (1520'F) 73.2

Estimate of 800K area (4 primary exhaust pipes): 0.10 m2

Radiated heat: 2 kW

Estimate of 800K area (muffler, turbo): 0.15 m2

Radiated heat: 3 kW

Say another 5 kW to be removed from under the cowling.

Total (39 * 1.22) + 5  = 53 kW.

Estimated exhaust gas energy 278 – 53= 225 Kw - that why a turbo is best for raw power

Assumptions

Similar efficiency for direct and indirect - via liquid - air cooling:

Inlet areas :

Cooling inlet radiator 210 cm^2 (estimated)

Cylinder barrel fins 20 cm^2 (estimate on diameter 5.0 cm)

Gear door louvers 18 * 7.5 * 1 = 135 cm^2 (18 louvers total)

11 ml gasoline per second with mass 11 * 0.7 g/s = 8.7 g/s

requires about 14 * 8.7 = 121 g/s of air per second

Assumptions

Air velocity into ducts is 40% of aircraft velocity

Specific Gravity 1.25kg/m3 at sea level pressure

Climb best rate = 120 km/hr (33 m/s) worst case air flow

Inlet radiator duct 210 cm/sq

Inlet supercharger 44 cm/sq

Specific heat capacity air = 1005 Kj/kg/deg K

Specific heat capacity water – 4.2 Kj/kg/deg K

Flow water = 73 l/min or 1.2 l/sec

Calculations:
Air flow supercharger

Using Q = (a*v*0.4) * s.g. = g/sec

44 * 33 * 0.4 * 1.25 = 726 g/s

NACA 3'' od Duct [44 sq cms]  will supply enough air by a 6:1 margin

Air flow radiator inlet

210 * (33 * 0.4) = 2370 l/s climb

210 * (60 * 0.4) = 5040 l/s cruse

Assume required radiator delta t = 10 deg /c (283 deg /K)

The Specific Heat formula is:  c = ΔQ / (m × ΔT)

Where:
  c: Specific Heat , in J/(kg.K)
  ΔQ: Heat required for the temperature change, in J
  ΔT: Temperature change, in K
  m: Mass of the object, in kg

Theoretical capacity air

1.005 = Q/ (2.4 * 1.25) * 10

      Q = 30 KW climb

dT = (Q / (Heat output turbo x Supercharger factor)) * target radiator
     = (30 / 39 x 1.22) * 10 = 6.2 deg C

1.005 = Q/ (5.0 x 1.25) * 10

      Q = 63 kW cruse

dTmax = 63 / 48 x 10 = 13.1 deg C

Theoretical capacity water

4.2 = Q/ (1.2 * 1) * 10  
Q = 50 kW

Rudder Cables - Part 3

This blog cover the installation/alignment of the rudder and associated cables

Turnbuckle

It was decided to use  AN155S Barrels / AN161 22S - AN161 22L  forks matching the overall length of the factory supplied S/S turnbuckles.

Push Rods
The original factory front ball joints supplied for the nose leg were found not durable in service and were replaced with plain rod end bearings . A 5 mm thick tube spacer was required to allow the rod end to clear the front attachment at full deflection. 

Note: The spacer was required to duplicate the original ball joint geometry and are now supplied with the kit

Stainless steel bolts were manufactured to ensure that only shank bare on all rod ends when installed with an  area washer fitted to prevent the ball joint escaping in the event of complete mechanical failure.

System Adjustment
The rudder fin was locked into its neutral position by using two aluminum channels fixed to the stab with automotive trim tape and foam wedges to lock it position.

The co-pilot pedals were locked in position with lengths of 16 mm curtain rod inserted into the pedals mounting allowing the cable to be adjusted and locking the rudder into its correct position. With the initial adjustment completed and the cables having a light tension, both aluminium channels were removed.

Note: A series of holes were drilled to achieve the correct location of the pedals in combination with packers at the end of the rod, wedged against the bulkhead. 

Sharpened bolts were used for quick install and release of the assembly during setup.

Pedal locks
White tube is 16 mm curtain rod

The pilot's pedals are fixed by the connecting rods relative to the copilots pedals, checking revealed that the previous work to set the centers was correct and both pedals are in a near vertical position with the cranks adjusted to the vertical position.

To align the nose gear a length 1'' x 1/4'' aluminum bar was mounted parallel to the nose wheel attachment and adjusted until parallel to the main spar using a measuring tape [not fun one up]. 

 

Nose wheel alignment gauge

The connecting rods that had been fabricated in house allowed only a few millimeters of adjustment so it was decided to remove both rods - shorten them by 12 mm, weld, strip, paint, primed and then repainted.

The cranks position were adjusted with the nose wheel location set as described above.

Pedal-Crank reference
marked for ease of
reassembly

I had chosen to drill the crank/pedals at this point and drilling in situ had disaster written all over it. 

With the the pedals and cranks aligned the position of the vertical bolt hole was marked through the crank onto the pedal shaft with a scriber. 

The pedals were removed, drilled and after dressing, Loctite no-sieze was applied and the pedals reassembled.

Both connecting rods were attached, bolts torqued as required then lock-wired.

Rudder wiggled
by the pedals

Note: Thread lock had been applied to the rod end retaining bolts on the previous installation, not a good idea as it penetrated the rod end and created a lot of extra time removing those bolts.

System movement was checked for free movement by hand - all seemed OK.

The cable will be tensioned once I can source a cable tension meter then lockwired to the video before closing the skin. Most likely the nose wheel will require a final tweak before testing.

Comment
Preferred aircraft graded turnbuckles.

The system is functionally complete requiring only final adjustments and lockwiring of turnbuckle

Merry Christmas 2016

Were are we now?

Currently I am moving from front to rear completing all areas to be ready for that final check and adjustment. The wings are nearly kitted out awaiting installation of the tanks and  finally the skins. A final task to be finished is the fuel connections to the external tanks which have yet to be finalized by myself and the factory.

In July I began to experience extreme pain in my ankles and this was put down to old age and working on floors at 2 deg C. After the boys from the factory left is got worse but work on I did with ever decreasing output.

Finally went to a doctor and after the usual tests it was diagnosed the membrane on the underside of the foot was damaged annoying the nerves and resulting in inflammation at the ankles. Cure is rest, meaning no power walking of any kind for the next six months plus a month on anti-inflammatory's.

Thank God for miracle of modern medicine as good health it the most precious commodity next to time in the world.

Next Year?
The goal is to finish the fuselage to the nose, close the wing and paint the air frame.

Merry Christmas to all

Battery Installation

This blog cover the relocation of the battery to the baggage compartment.

Overview
The factory supplied spreadsheet showed a real need to relocate mass to the rear of the aircraft. Currently the factory installs a 5 kg plus battery into a compartment located behind the baggage compartment rear wall and is particularly important for the supercharger installation. It was not practical to install the necessary structure as this needs to be done when the access is available before fixing the turtle deck, so it was decided to locate the battery on the inside the baggage compartment, enclosed by a battery box. 

Another reason is one of the aircraft on the field was a RV8 with the battery in the back fitted onto the fuselage floor behind the back seat - I was advise not place it there as acess is a bear.

Manufacture
The first task was to provide a means for the battery to vent any fumes to the outside of the fuselage. It is envisioned that the door at the front will provide some isolation but a sealed battery box was seen as a necessary item to prevent spread of any fire.

A length of louver was installed in the side panel of the battery compartment.

Battery Vent

A 
A 1.0 mm thick aluminium sheet a box was fabricated to provide a sealed cover over the battery pack open to the the vent shown above. To increase the fire rating of the aluminium the interior was lined with fire blanket cut to fit using a 60 mm diameter Ofla  rotary cutter. A cardboard template was cut, trial fitted and the profile traced with a soft pencil onto the blanket and cut.

The blanket was fixed using high temperature silicone by applying small dabs of silicone and then fixing the blanket. All the seams were coated with a bead of silicone to seal the edge and prevent the blanket fraying. 

Trail fit of blanket
The returns to the flange were removed

The floor had a blanket rectangle cut and covered with 0.016'' aluminium with the edges sealed to the aluminium with silicone and then fixed to the floor under the cover to protect the deck.

Battery
The goal of all this work was to relocate the battery mass from the firewall to the rear of the aircraft. This requires the installation of the two [2] battery's and 4 gauge cables to the firewall.

To save weight aluminium/copper coated multi core cable was selected and sourced from Perihelion Design ,this cable combines the performance of copper wire with the weight advantage aluminium saving 2/3 of the weight of the copper wire.

Note: The weight saved is about 300 grams [0.75 lbs] - hey it all helps

To allow connection of the cables a 25 x 3 mm aluminium bus was fabricated then covered with heat shrink, trimmed providing access to the studs and bolts and fixed with M6 bolts fitted with spring washers. 

Battery installation

The battery's now forms a single block making handling easier.

To fix the battery's M5 rivet nuts were installed into the deck and two [2] length of M5 S/S threaded rod cut and installed. A clamp was fabricated using 25 x 10 aluminium channel extrusion, alodined and then fitted with M5 lock-nuts.

The cable entry's points were marked using a stud with a pencil inserted to transfer the location to the compartment floor. Two holes were drilled to accept IP65 glands to clamp the cables and seal the compartment in the event of a battery failure while supporting  the cables when unbolted from the bus assisting in battery installation.

Cable Installation
The biggest task was to find a route for the cable remembering my aircraft is unique to the factory aircraft. The new manual details the installation of items like fuel and these BIG cables but this was not available when construction begun so this installation is unique. 

Moto: Follow the factory book of words

This installation will be detailed in a separate blog once the factory approves the route past the rear spar.

Comment
Follow the moto

The battery pack weighs about 1.0 kg versus the factory pack of 5 kg requiring a mass to be fixed the stab, calculations show this combination is about 1.5 kg lighter overall.

Starting again the battery would be fitted to the factory specification and additionally look at installing another access hatch on the opposite side and look at fitting the hydraulic pump on that side. This is common practice with a lot of factory aircraft as the pump is a large movable weight with little weight penalty connecting it back to the manifold. 

The key to the c.o.g. is if the nose wheel weight, if it goes over 88 kg you will need a larger mass at the fin. Right now all bets are off and every concept and piece of hardware is up for review and weight reduction.

Bits & Pieces - No 6

This blog covers the small tasks to finish the elevator installation

Rudder / Elevator Stab Fairing
After the disaster of nicking the spar plus the well documented repair it was decided this was to be a slow and careful installation. 

The first task was to mark out the locations of the ribs and spar to ensure that the drilling advanced the process not create more unproductive work.

Marking out rib locations before drilling

The top holes on the port side did not create any issues so there location was now carefully reproduced on the starboard side - then checked.

All holes are fitted with M3 rivet nuts and the fairing drilled 4 mm to provide additional clearance for ease of fitting. 

All unused holes were filled with spot putty, sanded and primed.

Rivets to fix the fairing allowing quick removal

On the underside only the front and rear fixing were installed and obtained a relatively neat fit with the lower stab.

Comment
Took another day to tidy up all these small loose ends in this area but it is now ready for paint preparation.  Should have done this the first time and would have avoided a lot of unnecessary pain.

Main Leg Lubrication

This blog cover the installation of greasing points into the main undercarriage leg

Overview
This builder decided to add greasing points to all high load pivots of the undercarriage for lubrication as required at bi-annuals /100 hourly.

Trailing Links

Bolt held in block, trimmed
then machined to length

All bolts were renewed to provide 100% bearing on links this required the purchase of a set of M6 x 90 mm Unbrako S/S bolts and remixing the factory supplied bolts. 

At installation M6 S/S spring washers were installed under each bolt head, this washer locks the bolt and arm together ensuring that all rotation occurs in the major attachment point located on the main gear leg. 

All bolts shanks were greased before installation.

Note: The changing of the bolts ensured that thread was not located in the working portion of components - shank support - thread retain. 

The center pivot blot was replaced with a M6 hex head to provide clearance to grease bolt on the main pivot mount. This was created by an unforeseen issue with clearance to the grease nipple on the main when the gear closes. [see note below]

Main Pivots
The boss face was drilled 3 mm through then drilled, tapped M5 to fit the grease nipple after facing the boss depth to 15.5 mm. To contain the grease at the rear a 1.5 mm cover was cut from 2024 aluminium sheet using a hole saw with the pilot hole filled using a 5 mm pop rivet and washer. 

Finally the cover was sealed with Permertex sealant, fixed onto the boss face using four [4] x M3 screws fitted into blind tapped holes. 

Note: After breaking a tap it was found 3:1 oil worked best as a tapping lubricant in this aluminium. A ratchet T wrench was purchased and proved to be perfect for this task.

Trial installation main pivots

The pivot arm was next filled using two [2] plugs of 25 mm thick builders foam foam cut with a ceramic hole saw. These plugs were pushed into the shaft and sealed with Permertex high temperature silicone.

Main Pivot Seal

Note: The hole saw was one in a bulk lot purchased to cut the holes for the seat belts with one perfect to produce the plugs. Plugs were fitted to prevent filling the arm with grease.

Hydraulic Cylinder Pivot
The rear of the cylinder mount was tapped M5 and a nipple fitted with a 12 mm hole drilled through the rib for access to the nipple. This position is not easy to access and will require a re-grease to be undertaken from the inside of the fuselage.

Note: Note sure about the access until the wings are mated

Drag Links

1/8'' Zerk drive nipple

Both drag link bolts are converted into grease bolts by drilling a 2 mm hole from the head side to the shank center with the head side redrilled to 1/8'' to accept the drive nipples, then finally a 1.5 mm hole was drilled to the pick up the end of the 2 mm hole and chamfered using a drill.

The M6 bolt hex was filled with silver solder and machined and drilled as outlined above with a nipple fitted. The disadvantage being it has to be tightened by clamping the head.

Cylinder is fitted with a stop collar to fix the stop point
stopping the gear just short of the top skin by about 5 mm

Stop Collar

The main M8 pivot bolt was drilled again as outlined above but it was found to foul the leg necessitating the machining of the head to a overall depth of 5 mm and the nipple reinstalled. The nipple overall height was reduced by 1 mm to improve overall clearance.

As noted above the center bolt in the trailing link group passed the nipple resulting in the cap head bolt beign replaced with a Grade 8.8 hex bolt - Murphy's Law in full application.

Another step forward

Comment
Doing it again the 1/16'' cover would be machined from 60 mm round and fitted with counter-sunk screws as the M5 rivet head made reinstallation tight.

All points where pressure lubricated and as expected more grease can be applied under pressure than by hand.

The limit switches will be covered in detail in a separate blog 

Bits & Pieces - No 5

This blog covers most of the last small tasks related to the rudder at final assembly.


Rudder Fairing

Tucano Tail

While only a just a little untidy it qualified for a upgrade and it's not a lot of work. 

A strip of 0.020" 2024-T3 sheet was cut, lower rivets removed, strip clecoed into position and rudder movement checked OK.

Once satisfied it was installed and riveted.

Note: The fairing bell mouthed slightly as it 
is really a developed shape and mine was straight but the result was acceptable.

Fairing upgrade

D Shackle

Rudder D Shackles
The factory supplied connection was replaced with a D shackles fixed with a clevis pin and castellated nut. 

This was changed as the original connection did not create a double shear connection but has now been upgraded for the aerobatic kits.




Rudder Adjustment
The top bearing is adjustable laterally to align the rudder mass balance and stab and is has slotted mounts for this task - after nipping one top bolt a blade screwdriver used to lever the rudder into its final location. 

There are two 3.2 mm holes provided to allow a permanent location relative to the vertical  spar.   

Fuselage Gap Seal
The junction between the fuselage and vertical spar has a noticeable gap this was sealer with Selleys Black SolarFlex silicone sealant. Two masking tape verticals wers applied and a bead of silicone added. This was then wiped using a disposable towel and allowed to cure.

First application of sealant
Tape removed after wiping silicone

After curing a second application was applied repeating the technique outline above.

Electrical Connections
The aircraft has been fitted with a white tail  light and red strobe. As an ultralight it cannot be flown at night but as it is in full camouflage it was decided to make myself a little more visible.

The connection was finished using the factory access panel to complete the connection. The plug is mounted on a tang located on the second rear bulkhead. the connecting cable in the fairing has a service loop to allow it to be extracted for servicing as shown below.

Final assembly before electrical testing / painting
The rivets shafts are a perfect match for the rivet nuts bore
and providing a perfect temporary fixing

Comment
Took a long day to tidy up all these small loose ends and others but it is now ready for paint preparation. Once painted the clevis bolts will be fitted, pinned as I want the rudder to be able to be deflected for painting.