ACM or Turbo Cooler

One of the things that set the L39 apart for owners is the quality of the air conditioning system. It is good and it makes flying the bird pleasant.

Moving from the AI-25 to the TFE731 series of engines presented some problems. Most of the problems I attacked head on and, to date, arrived at solutions that have worked incredibly well. The air path management and particularly the post fan exit path attention to detail has provided for an installation that is very efficient, simple and cost effective. The Air Cycle Machine (ACM) or, as its know in the L39 world, the turbo cooler was one area where I kinda stuck my head in the sand and hoped for the best. I regret that decision.

The original AI-25 has but one bleed air source. Subsequent testing has shown N1/PSI/DegF for AI-25 bleed air is Idle 10/260, start of cold air 68/15/240, cruise 99/75-80/350 and take off 107/105/430. The L39 uses a flow regulator in the ACM bleed air supply line (and one in the inlet de-ice bleed air supply line but the de-ice flow is WAY more than the ACM flow) which controls the maximum volume of air provided to the cooling and cabin pressurization system. Testing has shown that the AC vents will blow ever harder (more air volume) all the way up to ACM inlet pressures (measured just before the ACM system air cleaner - that big chrome thing in the engine bay) of about 22 psi. Beyond 22 psi, the flow rate to the cockpit does not change very much. The L39 uses a ball flow restriction type valve. The easiest way to imagine this valve is to make a horizontal circle using your thumb and forefinger. Now imagine that you have a ping pong ball sitting on that circle forming a valve and, given the ball is resting on your fingers, it is closed. If you try to blow air down past the ball and through the circle you made with your fingers, none would pass. Now imagine a spring holding that ball just off your fingers so air can now flow. The more air that flows around the ball the more that air will push the ball down against the spring and finally to your fingers. This is the fundamental mechanism used by the L39 flow control valves. The spring is set such that the desired flow is achieved. Attempt to flow more and the valve closes; flow less and the valve opens. The regulation point on Grey occurs at about 22 psi input pressure and it is at this pressure and above where X amount of Kg/sec of air flow is achieved.

Aero Vodochody installed a Garrett TFE731-4 motor in the L39 chassis and called it the L139. The TFE731 series of engines have a high pressure (HP) and low pressure (LP) bleed arrangement. Business jet applications use mixing valves that combine HP and LP bleed sources into a common ACM supply. The WestWind valve, for example, relies on HP bleed until manifold pressure gets to 18 psi at which point the HP poppet valve starts to close. By 22 psi (depending upon the series of valve), the HP poppet is completely closed and the ACM system is supplied by LP bleed air only. The relationship between HP and LP bleed on a 731-3 variant is shown here. The 731-3's LP bleed is approximately 66% of the HP bleed's pressure at lower N1s. It does not take long at low altitudes for the bleed mixing valve to transition to the LP source. If my data is accurate, it happens about 50% N1 or basically anything above idle. Normal business jet applications will transition back to HP bleed air at much higher altitudes as LP bleed pressure falls but the L39 is not RVSM compliant and thus never sees these altitudes.

I've not verified these comments but I have heard that most TFE731-3 installations in L39s have very poor AC performance on the ground. This would make perfect sense if these installations relied on LP bleed air only for the AC system as 731-3 HP bleed air is only 9 psi at idle and the L39's ACM does not start blowing cold until about 15 psi (again, measured at the entrance to the ACM system air cleaner). Given this, I was faced with designing a mixing system like AV did on the L139 or sticking my head in the sand and using HP bleed only. I stuck my head in the sand and that was not wise. The ACM expired in a dramatic way during our first cross country simulation to FL25. The turbo shaft over sped causing the compressor wheel to shatter.

It is worth pausing here to understand exactly what drove the failure and why differences in the AI-25 and TFE731 contributed to the cause. The AI-25 is not terribly efficient. This statement is born out by our initial cruise data at FL25 showing a 33% reduction in fuel usage in high speed cruise at exactly the same speed (AI-25 at 150 gph versus the 731-3 at 99 gph both at 350 TAS). This difference in efficiency shows up in the bleed air extraction. I have yet to get recorded data on LP bleed pressures and temperatures but it is more than likely that the AI-25s final stage bleed air profile more closely resembles that of the 731-3's LP bleed. Given that the L39 ACM flow restrictor sets bleed extraction to X amount, the only difference between using AI-25 final stage bleed air and 731-3 HP bleed air to supply that X amount of flow is the temperature energy of that mass flow. The 731-3's HP bleed air is much hotter. Simply put, the ACM uses heat energy extracted in the expansion turbine portion of what amounts to a simple turbo charger type device to drive the compressor. The compressor is used as a fan to move air across a heat exchanger. Put too much heat energy into the ACM system's X amount of mass flow and the ACM itself is stressed. Further take that ACM to the FLs which presents thinner air and thus much less loading on the compressor wheel and, viola, the shaft over speeds and the compressor comes apart from centrifugal forces. Had I taken time to fully understand the system, results would have been predictable.

We are now in the process of installing a WestWind bleed mixing valve on Grey for testing. I'm also designing and will soon test an engine resident mixing system that is designed to (1) be more in line with L39 flow requirements and (2) will not be based on $15K certified aircraft mixing valves. The L39s ACM and cabin pressure system's bleed air volume requirements are much lower then those of the business jets these engines and bleed mixing systems come from. I say this not from physical flow measurements or aircraft documentation but from anecdotal evidence of how the systems perform and their affect on TIT when employed. It is obvious that the cooling and pressurization needs of a Hawker are much greater than that of an L39. The SOP for the Hawker is to take off with bleeds off to maximize power. Once established in the climb, bleeds are turned on. This normally results in an approximately 20 degC rise in TIT as bleed air energy is robbed from the engine's core. Conversely, taking off with the L39's HP bleed off with the throttle set to 885 degC (max continuous cruise power setting) has TIT falling by about four or five degrees by the time you reach for the gear handle to pull up the gear. Turning ACM bleed on at that point brings TIT back to the original 885 degC. Put differently, the penalty for Hawker LP bleed to ACM usage is four times more than the L39's HP bleed usage. Clearly the L39 uses less than a 1/4 of the volume used by the Hawker. The bleed mixing system for the L39 need be nowhere near the size/volume of the Hawker (or WestWind for that matter). In addition, these conversions and support costs must be kept to a minimum to provide the most value to the fleet of flying L39s. Not using expensive overkill mixing valves is one way to control costs.

My goal is a solution that uses a bespoke mixing manifold bolted to the LP bleed port and incorporating the backflow valve to the engine's LP supply (to keep engine HP bleed from back flowing to the LP port when HP is turned on). The HP supply to this manifold can come from the top HP bleed exit as volume is low and HP bleed is only required at or near idle speeds. A small electrically operated pneumatically assisted valve is used to control HP bleed use with an electrical control that manages HP bleed usage based on pressure with an over temperature function. This will mirror the system AV employed on the L139. I also want to carefully consider inlet de-ice requirements to make sure there is sufficient heat energy in the de-ice bleed air when descending from the FLs using idle power.

We now have a HP/LP bleed mixing valve installed from a WestWind. The supply temperatures and pressures to the ACM system for the TFE731 are now looking much more in line with those from the AI-25. Here is a quick systems check. Although TO power was not held for an extended amount of time, the maximum system inlet temperature was right around 410 degF which compares favorably to the AI-25's 430 degF. Maximum system pressure was down from 105 psi for the AI-25 to 60 psi for the TFE731-3. Keep in mind that the ACM system's flow control valve only needs about 22 psi of inlet pressure to the system to provide all required cooling and cabin pressurization needs. The next step is flight testing the system.

In parallel with installing the WestWind bleed mixing valve, we've been developing our own mixing manifold as mentioned above. The first article will be completed in a few days and then installed on an engine for in service testing. Below are some renderings. The HP/LP mixing portion is integrated in a manifold that bolts directly to the TFE731-3's LP bleed port. A pneumatically controlled valve is added to feed HP bleed air into the top of the manifold and the combined HP/LP mixed bleed air exits the bottom of the manifold to both the Turbo Cooler and inlet de-ice. This is almost identical to the system used on the L-139. Once combined manifold exit pressure reaches around 22 psi, the HP valve starts to shut. Once LP air exceeds 22 psi, the HP valve is completely closed. This allows good ACM performance on the ground at idle by relying on HP bleed air then transitioning to LP bleed air in flight. The Turbo Cooler is not stressed and the best efficiency is achieved by using LP bleed air in flight.