Carbon Ball Blasting
Carbon ball blasting is caused by particles of fuel that have not been completely consumed. It usually the result of fuel particles being too large to be consumed. The size of the fuel particles after being atomized is referred as the Sauter Mean Diameter. As the fuel nozzle gathers operating hours, the fuel particle Sauter Mean Diameter can be larger in emitting from some fuel nozzles and not in others. Hot refueling often results in very fine sand particles getting into the fuel and the particles are too fine to be caught by filtration. Often the sand will become packed on the screen inside the fuel nozzle which affects the atomization of the fuel.. Often the sand causes erosion of the slots in the nozzle orifices and the end result is larger fuel particles. Larger fuel particles are not completely consumed and the particles are in the form of carbon. When the blade tip is above approximately 1650 degrees F, the rate of erosion from the carbon balls increases. When possible, an operator should consider sending the fuel manifold to a Service Center for flowing, cleaning and possible fuel nozzle replacement. Usually at approximately 600 hours, the cleaning is effective in restoring the fuel nozzles to their proper flows.
The inner wall section of the 1-130-780-01 combustor liner part number is 1-130-253-01. The 1-130-253-01 section of inner wall has dimples that are 0.030 inch in depth. They offset the inner wall 0.030 inch from the nozzle curl for cooling air that follows the wall of the turbine nozzle curl and cools the top 1/3 of the Gas Producer turbine blades. The nozzle curl where the liner dimples contact it, wears and as it wears, the cooling air is reduced. The nozzle curl can be welded and then ground to conform to the outer circumference but the 1-130-253-01 inner wall should be replaced. IT IS A MISTAKE TO BEND THE TABS OF THE Liner against the nozzle curl because the rate of carbon ball blasting increases and erodes the turbine blades. However if the tabs do not contact the curl, they may eventually break off. (From vibratory stress)
The 1-110-500-03 deflector is supposed to have a chamfer as a lead into the combustor liner outer flange. When installed in the engine, there should be a clearance between the outer flange on the combustor liner (where both parts contact in operation) and the 1-110-500-03 deflector to allow for thermal expansion of the liner. The clearance is approximately 0.050 to 0.070 inch. If the gap is excessive, the engine will vibrate and RUMBLE at idle. If there is little or no gap, the 1-110-500-03 deflector will crack around the inner flange where it is bolted to the air diffuser. When a deflector is repaired, caution has to be taken to control the dimension between the mounting flange of the deflector and the contact area where it meets with the combustor liner. Global now controls that dimension.
The advantage of an annular combustor is that a radial temperature profile from base to tip can be obtained quite easily and will benefit the life of the turbine blades. A more compact engine envelope is also obtained. The 44 chutes in the outer wall provide cooling air to the base of the 1st stage GP blades. In the T53-L-13, 13A and 13B, the base of the blade is actually 150 to 170 degrees cooler than the gas temperature of 1730 degrees F at Takeoff power. Two thirds of the way outward on the blades, the blade leading edge temperature is equal to the gas temperature of 1730 degrees F. IF the inner wall is offset by 0.030 inch dimples from the nozzle curl (by design), the blade leading edge temperature will drop a minimum of 75 degrees from the hottest point of 1730 degrees F (2/3 outward from the base) to the tip.
By having the hottest point of the blade 2/3 outward from the base predetermines that should a stress rupture occur, it will occur 2/3 out on the blade and not at the base. It is what I consider a "soft" failure because the engine should continue to run and make adequate power.
The stress rupture life of the blades decreases as the metal temperature increases. The stress rupture life (in the Design Report) was based on a radial temperature profile of 1580 degrees F at the base increasing to 1730 degrees F at the leading edge 2/3 of the blade length outward from the base and then decreasing to 1655 degrees F at the tip. The gas temperature at 1730 degrees F turbine Inlet Temperature drops across each set of turbines. A good performing engine will have approximately a 700 degree F temperature drop across both sets of turbines. Approximately 61% of the temperature drop is across the Gas Producer Turbines and 39% across the free power turbines. By set of turbines, the turbine nozzles are included.
It is not the same for all engines because the temperature drop varies as the nozzle areas vary. A good performing engine will have a temperature drop of approximately 420 degrees across the GP turbines. Approximately 2/3 of that occurs in the 1st stage GP Turbine. That means the gas temperature exiting from the 1st stage GP turbine has decreased from 1730 degrees F to approximately 1450 degrees F. That temperature gradient is approximately the same as the metal temperature gradient across the blade.
The T53-L-703 is subject to excessive carbon ball blasting. The reason is not clearly understood because the 1st GP turbine blades are cooled by compressor discharge air and should have a lower metal temperature than the blades in the T53-L-13B. A comparison of the total area allowed for cooling of the blade tips leaves doubt. Some of the holes for cooling air did not consider that the holes around the circumference are partially blocked when the 1-130-780-03 is inserted onto the nozzle curl. I believe that the T53-L-703 tip cooling area is approximately 60% of the area of the design requirement of the T53-L-13B.
Air Technology Engines, Inc. is considering the submittal of an Engineering Order to the FAA that would allow additional holes in the forward section of the 1-130-780-03 combustor liner. The intent is to increase the amount of cooling air to the tips of the turbine blades making the area in the T53-L-703 (T5317 models) the same as the T53-L-13B (T5313B).
The atomization of the fuel is dependent on fuel quality and cleanliness. Fine particles of sand, small enough to get through the aircraft fuel filter, get lodged against the primary and secondary screens in the fuel nozzles and affect the amount and quality of the fuel going through the 0.004 inch orifices into the swirl chambers. There are two different fuel sprays (primary and secondary) emitting at different angles and creating a common spray angle directly in front of the nose of each fuel nozzle. Proper atomization is dependent on the proper flows coming from the primary and secondary fuel orifices in the fuel nozzles into the swirl chambers. As the flow falls off due to contamination, the particles of fuel increase in diameter and become carbon balls as a result of incomplete burning. The quality of the fuel and proper refueling procedures can reduce the carbon ball erosion problem.
It is important know your engine and how it performs. Cooler running engines last longer. When you know that the engine requires a higher EGT or TGT to make power, it is prudent to find out what is wrong. Even though the engine makes adequate power, more than likely costly damage is being done to the hot section of the engine. Sometimes all the engine needs is a compressor cleaning.
My name Is Edward C Pease (Ed Pease). I began working at Lycoming in 1957 as a test engineer on the T53_L-1 engine. In 1965, I was in charge of the development of the T53-L-13. For a period of four years I managed the development programs of all of the turboprop and turboshaft T53 engine models. I took early retirement in 1987 and became a turbine engine consultant on many makes and models of turbine engines. After 60 years in turbine engines, I limit my consulting to two overhaul shops in Florida. My assistant (John E Moore) at Lycoming for many years, still works with me at Air Technology Engines, Inc. We do their engineering which includes the development of repairs.
Originally Posted 10/16/2016
