Lockheed Model 89 Turbosupercharger Selection
Turbosupercharger Design for a Transport Airplane
Compiled by Kimble D. McCutcheon
Published 17 Jun 2025

14 Sep 1943. Phillip A. Colman released Lockheed Report No. 4360, Turbosupercharger Design for a Transport Airplane, which described part of an airplane design intended for military and commercial operation, predicated on high speed and long range. Based on power required, engine size and engine weight, the Pratt & Whitney XR-4360 was selected. The advantages gained from high-altitude cruising flight were many and assumed an exhaust-driven supercharger whose characteristics matched the XR-4360 single-speed single-stage engine.

Design Considerations

Lockheed Report No. 4342 outlined the turbosupercharger design requirements for a transport airplane, summarized below.

Altitude. The cruising speeds of an airplane for maximum-economy (maximum miles per gallon) to maximum power increase with altitude. Higher cruising altitude results in less time to travel a given distance. At 25,000 ft the time to fly a certain distance at maximum economy power is 28% less than at 10,000 ft. Average meteorological conditions improve rapidly at altitudes above 18,000 ft; the frequency and intensity of stormy conditions diminish. Temperature and humidity conditions conducive to airframe icing are practically non-existent above 18,000 ft. Airlines benefit from added passenger comfort due to turbulence-free air and improved flight schedule regularity. In 1943, most airline operation was at 12,000 ft and lower; flights were often cancelled due to weather, which would have happened much less frequently if 25,000 ft cruising altitude had been available. High altitude operation resulted in minimal fuel consumption due to reduced exhaust backpressure; a reduction in backpressure from 30 inHgA to 25 inHgA reduced fuel consumption by about 1.7% with the engine operating at 140 psi bmep. This was equivalent to 810 lb of fuel for a 5,000 mile range with the proposed Lockheed Model 89.

Size. Nacelles on a four-engine airplane constitute a large portion of total airplane drag, which makes minimum nacelle frontal area important; a 6" nacelle diameter increase adds about 3% to total airplane parasite drag, which corresponds to a 1.5% increase in fuel required, or 720 lb (four passengers) on a 5,000 mile flight.

Weight. Every pound increase in empty weight subtracts about 1.3 lb from the payload; the 0.3 lb is the extra fuel required to carry the extra pound.

Takeoff and Climb. In order to operate from major airports worldwide, transport aircraft must be able to take off from altitudes above sea level. To operate efficiently at high weights, the sea-level rated takeoff power should be obtainable at the highest normally used airport. To take full advantage of high cruising altitudes an airplane must maintain high engine output to altitudes approaching the cruising altitude without sacrificing specific fuel consumption. Time to climb to cruising altitude and fuel consumed must be sufficiently small to take advantage of the high speed obtained at altitude.

Reliability. With its need for utmost reliability, transport airplane components must at least equal the service reliability of the basic engine.

Turbosuperchargers

Turbosuperchargers can achieve the manifold pressure required to maintain cruise power at high altitudes while only slightly increasing exhaust backpressure and thereby costing a negligible amount of additional fuel when compared to engine-drive superchargers. By careful design, the backpressure resulting from cruising powers at 25,000 ft can be less than sea level static pressure, resulting in a specific fuel consumption lower than at sea level. Since only cruising powers are required at 25,000 ft, the turbosupercharger may be relatively small and light. Since size, weight and fuel consumption intimately affect transport airplane efficiency and are closely bound to safety requirements, considerably more design ingenuity must be expended on turbosupercharger design than is necessary for military bomber and pursuit aircraft. The design effort to obtain maximum compressor and turbine efficiencies will be repaid many times in airline operation savings. Required nozzle box pressures must be as low as feasible to obtain maximum fuel efficiency. Turbosupercharger size, shape and adaptability must not unnecessarily dictate nacelle dimensions or arrangement. Weight must be the minimum consistent with system reliability.

Turbosupercharger Specifications

Cruising turbosupercharger specifications were presented in Lockheed Report No. 4342, with size and installation requirements explained in Part A and performance presented in Part B of this Report No. 4360. Operating in conjunction with the Pratt & Whitney single-speed, single-stage XR-4360 engine, with intercoolers, heat exchangers and other components described in Part A of this report, the following minimum powers and altitudes shall be obtained:

Comparison to Existing Turbosuperchargers

Part A. Turbosupercharger Installation and Design

A new turbosupercharger design should be simply and effectively incorporated into a transport airplane nacelle. Lockheed, which studied such a design to determine an optimal shape and arrangement, concluded that an axial compressor driven by a modified GE Model B turbine would be ideal. This would fit in an envelope that was 40" long (including entrance spinner and turbine cooling cap), 29" wide (including exhaust inlets) and 15" high.

Lockheed studied many component arrangements, evaluating their effects on aircraft aerodynamic efficiency, weight, adherence to standard construction techniques and maintenance practices, and power plant efficiency and reliability. Lockheed chose the Pratt & Whitney R-4360 engine in a standard nacelle whose frontal area would be unaffected by the turbosupercharger installation.

Fig. A-1. Induction System

Induction System. Figure A-1 shows the desired turbosupercharger installation. The engine air path was as direct as possible between the forward-facing ram air scoop to the carburetor at the engine back. A straight passage through the compressor with double outlets provided an efficient duct system from the scoop to the intercoolers, which were positioned so that straight ducts were installed for the cooling air without excessive engine air ducting. Since a carefully ducted and vaned 90° turn resulted in a dynamic pressure loss of 15% to 25%, the straighter the duct system the greater the overall power plant efficiency. This efficiency was directly related to the turbosupercharger work requirements, which reflect in the maximum system altitude performance and required exhaust back pressures.

Double turbine inlets (Fig. A-1) lessened the exhaust system collector complication and eliminated sharp exhaust pipe bends. The airplane design conditions required an exhaust heat exchanger for wing and tail de-icing, which complicated the eduction system. A study of installation and turbine operation problems indicated that heat exchanger location aft of the turbosupercharger to be better. This allowed a shorter overall length and avoided transition sections, reducing exhaust backpressure. The use of a single large heat exchanger eliminated the complication of split systems. The straight heat exchanger exit pipe allowed the application of a reduced nozzle exit, obtaining the maximum thrust energy recovery consistent with engine backpressure requirements.

Rammed air was required for engine cooling, oil cooling, intercooling, turbosupercharger cooling, deicing, cabin heat, and to feed the carburetor. A single ram air scoop was most efficient, but this required the air-consuming units to ideally be in line with the air intake. The nacelle arrangement depicted in Figure A-1 accomplished this. The nacelle's overall frontal area was only 25 ft², smaller than many others housing less powerful engines without turbosuperchargers, intercoolers or heat exchangers. One reason for this is that the retracted landing gear was not housed in the nacelle and the engine installation alone determined the nacelle dimensions.

This nacelle design also promoted in-flight power plant accessibility; locating the turbine unit in the nacelle bottom not only allowed access to the turbosupercharger, but also the engine and accessories.

Fig. A-2. LTS End, Side and Plan Views

The shape and size of the LTS is shown in Figure A-2. Lockheed studied the attachments and connections to optimize installation. Instead of the normal flange connections at the compressor outlet, it was beaded to accommodate flexible members. The nozzle box inlet also eliminated flanges and substituted a cylindrical pipe section for a standard exhaust expansion joint.

The compressor was a 7-stage axial blower with its rotor divided into two sections. The 4-stage entrance section was driven directly from the turbine wheel shaft. The 3-stage aft section was driven through a hydraulic coupling connected to the turbine wheel shaft. A suitable speed control regulated the aft-section speed to obtain a wide range of operating conditions.

The compressor nose casting provided a flat annulus for an entrance air duct seal and a series of vanes directed the incoming air to the main rotor first stage at the most efficient angle. The nose casting housed the fluid coupling controls and main rotor front bearings. These were accessed through a removable spinner. Both the entrance and aft rotors were castings with multiple cast blades that were adjustable and removable. Between each rotor stage were stationary vanes connected in sets to support rings and assembled as a unit in the main compressor castings. Within the rotors were a stationary oil tank, oil pump, fluid coupling, speed control and rotor bearings. The compressor outer case was made in two sections, cast with sufficient ribs to provide for air cooling along with nose casting and compressor discharge housing support.

The compressor discharge housing was a cast scroll with dual exits, attachment for the turbine nozzle casting, and supports for the main shaft center bearing. This center support was radially ribbed to give the required strength and extra cooling. Air was admitted through six cutouts in the housing between the ribs and passed over the internal cooling diaphragm portion of the nozzle casting. The turbine wheel was cooled by radiation.

The turbine nozzle box was corrosion resistant steel weldmemt with double exhaust inlet connections that were cylindrical for at least 2" to provide a suitable tail pipe expansion joint. The nozzle box was shaped to have a minimum outside diameter, obtaining the required internal area by increasing the fore and aft dimensions. The nozzles had a 16.1 in² effective area. The nozzle diaphragm was a steel casting with suitably cambered turbine nozzle vanes. Its inner portion was radially ribbed to cool the turbine wheel by radiation and support the main shaft aft bearings.

The waste gate was a sleeve within the nozzle box that surrounded the nozzle casting. This continuous ring slid axially through a 2" travel and was robustly made to act as a guard in case of turbine bucket failure. In the open position it just covered the turbine bucket outer edge; in the closed position it sealed against a flight hood adapter and provided a suitable outer surface for the turbine exit diffuser. Waste gate control was via a series of levers, connecting links and push rods. Two removable push rods were connected to the waste gate ring and extended to the outside through bearings provided in the flight hood adapter. Short links connected these push rods to levers that were mounted on a common shaft to provide synchronized motion of both waste gate sides. The flight hood adapter provided suitable bearings for this shaft and a sealed tube surrounded the shaft as it passed through the cooling cap. The shaft outer ends were grooved where the operating levers attached on either side. All bearings, attachments and moveable joints were designed with extra safety margins to ensure satisfactory operation and a long service life.

The flight hood adapter was a cast steel ring attached to the nozzle box rear edge via screws, allowing access to the waste gate. The adapter's forward inside edge was the seal line for the waste gate in the closed position. The adapter also provided support and bearings for the two waste gate operator push rods and bearings for the waste gate control shaft. A cylindrical section about 1.5" long was provided as part of the adapter and formed an expansion joint for a tail pipe between the turbosupercharger and heat exchanger. When the was gate was closed or nearly closed, the inside surface of this cylindrical adapter section also acted as part of the exhaust exit diffuser.

A cooling cap comprising an exhaust exit diffuser and radiation type cooling cap was installed on the exhaust turbine wheel aft side. The cooling cap was designed in accordance with General Electric Report No. 47103, Design of Exhaust Hoods for Turbosuperchargers. It was made from corrosion resistant steel sheet formed and welded into an assembly that provided an airtight radiation baffle over the exhaust wheel with suitable passages for cooling air circulation, along with inlet and outlet connections for the air supply. Air for the cooling cap was taken from the heat exchanger entrance duct, passed through the cooling cap housing, circulated over the turbine cooling diaphragm rear face, and exited to the hot air side of the heat exchanger duct. A pressure drop of approximately 6 inH₂O across the heat exchanger was available for air circulation through the cooling cap. The unit was supported by vanes that were integral with the adapter casting. The connections to the supports allowed the cooling cap to be removed or adjusted for proper turbine wheel clearance.

The turbine wheel was a modified GE Model B wheel. The basic wheel portion was used with redesigned buckets to obtain favorable nozzle box exhaust pressures. The wheel became a 50% reaction type with a 1.125" bucket length.

Part B. Turbosupercharger Performance Calculations

This report section expanded the data from Lockheed Report No. 4165, which presented compressor performance for design conditions only. The data were expanded to include turbine and compressor performance at all usable engine powers.

Fig. B-1	Fig. B-2

The 7-stage axial compressor was designed for maximum efficiency at engine cruising powers between 1,000 and 1,600 bhp at 25,000 ft, with the peak efficiency occurring at 1,300 bhp. Figure B-1 shows a plot of available engine power versus altitude and illustrates the maximum turbosupercharger capacity. A critical altitude of 20,000 ft was expected bhp at 2,500 engine bhp, resulting in a 2,160 bhp maximum power available at 25,000 ft. A minimum usable engine power of 900 bhp at 25,000 ft was desired to satisfy the lower gross weight airplane cruising requirements. Takeoff power was available to 1,500 ft fro the engine without turbosupercharging, and previous estimates based on compressor capacity indicated an increase in critical altitude to 13,000 ft when operating with the turbosupercharger. This Report No. 4360 showed that data and calculations involved in checking the turbine and compressor capacities at cruising, maximum and minimum engine power conditions.

Compressor Performance. The axial compressor working chart (Fig. B-2) was constructed around the pressure ratio versus volume flow curves given in Lockheed Report No. 4165, Turbosupercharger for the Wasp Major Engine, by J. Van den Bouwhuysen. These curves showed the compressor capacity at 25,000 ft and by applying an inlet temperature correction for varying altitudes to the pressure ratio scale, the curves were generalized so that compressor performance at any altitude could be determined. The addition of curve sets for graphical determination of the volume flow and pressure ratio for any engine power at any altitude makes it possible to enter the chart with engine air mass flow and compressor discharge pressure and proceed as indicated to find the compressor speed. (The given carburetor pressure was increased by the intercooler and duct pressure drops to find the required compressor discharge pressure.) This graphical determination was made for all critical engine power and altitude conditions, and the resulting line sets of constant engine power and constant pressure altitude are shown superimposed upon the compressor speed curves. Engine airflows, carburetor pressures and intercooler pressure drops necessary for the construction of the constant power and altitude lines are obtained from Figures B-5 and B-8. The performance picture the curves presented on the compressor working chart was complete insofar as the compressor was concerned. The location of the 900 bhp line with relation to the surge boundary indicated stable compressor operation at the minimum required engine power to 25,000 ft. The critical altitude for 3,000 bhp and 2,500 bhp were shown to be within the compressor capacity limitations when running at approximately 21,000 rpm.

Fig. B-5	Fig. B-8	Fig. B-3

Compressor Efficiency. Analysis of turbosupercharger performance requires knowing the compressor shaft efficiencies. Report No. 4165 gives efficiencies for use with a polytropic exponent of compression of 0.287, but since the use of adiabatic efficiencies is more universal in turbosupercharger work, the polytropic values were converted for use with the adiabatic exponent (0.283). A plot in Report No. 4165 relating the developed head (ft-lb/lb), the compressor speed and volume flow was given. From that plot the set of adiabatic shaft efficiency curves on Figure B-3 was made using the following expression for solving for Ŋ_T at respective values of the load coefficient Q/N:

shp = (H x M) / (33,000 x Ŋ_i) = (T x Y x Cp x M) / (Ŋ_T x 42.4) , where

shp = turbine shaft horsepower required,
H = developed head (ft-lb/lb),
M = Air mass flow through the compressor (lb/min),
Ŋ_i = Polytropic shaft efficiency
T = Absolute inlet temperature
Y = Adiabatic pressure ratio function = ((P1 / P2)^0.283 ) – 1,
Cp = Specific heat of air ~ (BTU/lb/°F),
Ŋ_T = Adiabatic shaft efficiency,
Q = Airflow volume through compressor (cfm),
N = Compressor speed (rpm).

Solving the expression for Ŋ_T = 779 x ((T x Y x Cp x Ŋ_i) / H)

The actual difference between the polytropic and adiabatic shaft efficiencies is not very great in the region of highest efficiency, but as efficiency falls of the adiabatic value decreases much more rapidly than the polytropic.

Compressor Discharge Temperatures. Compressor discharge temperature versus engine power curves at 25,000 ft are presented on Figure B-9. The Standard Day curve came from Report No. 4165 and the hot-day curve was computed using the Ŋ_T versus Q / N curves from Figure B-3. carburetor air temperatures arising from engine-turbine operation at 25,000 ft on both NACA Standard and Army hot days were checked in Report No. 4165 and could easily be maintained below the limiting values with the proposed intercooler installation. The critical altitude conditions for 3,000 bhp and 2,500 bhp appear to be critical and were checked in Table 1, which indicated that on a Standard day, both conditions could satisfactorily be handled, since 100°F was the upper limit on carburetor air temperature. On an Army hot day, the critical altitudes will be reduced and the carburetor air temperature would be more than 40°F higher at the same compressor speeds. If the turbine capacity were sufficient to carry 3,000 bhp to approximately 10,000 ft on a hot day, the carburetor air temperature was expected to exceed the 100°F limit.

Turbine Performance. In order to determine engine backpressures at critical engine power conditions, the turbine performance had to be estimated. A peak turbine wheel efficiency of 65% was assumed and a typical efficiency curve was drawn through this point as shown on Figure B-4. The resulting curve was used with the three General Electric curve sheets, Figures B-10, B-11 and B-12, to compute values for turbine speed, pressure ratio shaft horsepower, and altitude relationships shown on the turbine working chart upper half, Figure B-7.

Fig. B-4	Fig. B-6	Fig. B-7	Fig. B-9	Fig. B-10	Fig. B-11	Fig. B-12

Table 2. Turbine Working Chart Sample Calculation

The turbine working chart was completed by an exhaust gas temperature correction curve was copied from the GE data, which was added to its top, and the addition of the lower curve sets for complete graphical determination of the nozzle box pressure required for a given operating condition.

The nozzle box pressure was lower than the back pressure on the engine exhaust parts by the amount of the pressure loss in manifolds and collector ring leading to the turbine, but in this analysis, only the nozzle box pressures were shown. An exhaust gas temperature of 1,600°F was used in all cases. Table 3 shows the computation of nozzle box pressures required for operation at standard conditions at all powers at 25,000 ft and at the critical altitude for both 3,000 bhp and 2,500 bhp. The available exhaust power was then checked on Figure B-12 to ensure sufficient exhaust gas flow to drive the turbine. The Figure B-12 chart is entered with the nozzle box pressure and altitude equivalent of the backpressure on the turbine. The required weight flows were presented in Table 4 in percent of the total engine airflow. Note that the increase in weight flow due fuel addition to the air is omitted in figuring the total available airflow. The fuel flow was assumed to balance exhaust system gas leakage.

Conclusions

[U.S. National Archives Record Group 72, Entry 145, Box 3. Lockheed Transport Model 89 Report No. 4360.]

Send mail to with questions or comments about this web site.
This website depends on cookies to make it function. If you continue to browse, scroll, click or otherwise interact, you are providing implicit acknowledgement of and agreement to this.
Copyright © 2002-2026 Aircraft Engine Historical Society, Inc.