Table of Contents
2.2.5 Start abort and new start
When a start abort occurs, one can assume that a maximum component temperature, on the grounds of transient operation, has not been reached. In such cases, one must normally not reckon with a full start appropriate influence on lifetime.
If a start has to be repeated shortly after a start abort, one should pay attention to the special steps that are written down (by the OEM). Included in this is the drainage of possibly given fuel remains, in the lower combustor area. Such fuel remains can lead to extreme local over heating of the combustion area as well as the turbine. In small performance gas turbines, there is the additional danger of overspeed of the engine, because the control range is overstrained through the remaining fuel. This can bring on spontaneous rotor failure. The occurrence of locally limited overheating, around the perimeter of the gas duct, can lie on the wrong functioning of the nozzle system (e.g., injection distribution of the nozzle).
"Illustration 2.2-1and2": Because of differences in thermal expansion of the rotor and the casing even during standstill ( "Ill. 3.1.2.4-2") the rotor can get temporarily stuck (jam). The contact occurres by the rub of the rotor blade tips with the casing and/or the rotor spacers ( "Ill. 3.1.2.4-1") respectively the rotor drum with the guide vane. This situation is to be expected in the first stage of operation after the initial start up, as long as the clearances are not cut in. A temporarily stuck rotor after shut down, may not signify a permanent problem. If the handbook does not state otherwise, one is permitted to start only when the rotor is free again.
A further, possibly, superimposed effect is a rotor bow ( "Ill. 2.5-3"), as a consequence of an uneven temperature distribution in the rotor. This temperature distribution occurs because the warm air in the gas duct rises during stand still (A) leading to a higher temperature on the top side of the rotor. This brings about stronger thermal expansion and bends the rotor (B). Depending on the type of the gas turbine, (e.g. derivate or heavy frame), such a condition can first emerge after hours and can remain correspondingly long. If there is such a danger, the manufacturer indicates the necessary waiting time ( "Ill. 2.2-1 and 2") before the possibility of a new start. Should this interval of time not be maintained, heavy vibrations and severe rub can occur. There remains the danger of self- reinforcement: during rub, the rotor is heated up locally and bends itself even more. In an extreme case the rotor drum can be rubbed through and heavily overheated.
The distortion of the rotor can concentrate on a mismatch of the flange and /or the interference fit. In this case, it can come to a sudden drop of unbalance with a loud ‘bang’ during the run up of the engine by finding a new center (see "Example 2.2-1").
This event may more likely occur on rotors with a central tension rod ( "Ill. 2.1-7") than at those which are bolted or welded at the circumference.
"Illustration 2.2-3and4": Turbine discs are highly subject to stress, especially during start (Lit.2.2- 1). The above picture presents the cross-section through a typical integral turbine wheel of a small gas turbine with temperature and stress distribution in steady state operation. We find such integral rotors, called ‘Blisks’ in turbines and compressors. This definition is an acronym of ‘Bladed Disk’. The scetches above show on the left the temperature distribution and on the right the thermal stresses during stationary operation. During start, the mechanical stresses on the hub are particularly high due to the rapid heating up of the disc rim, via the blades, which generates a big temperature gradient (picture at the bottom) between the hub and rim after a few minutes. This is accompanied by correspondingly high thermal stresses that superimpose the centrifugal force conditioned, tensile stresses in the hub. The tensile stresses are thus further raised. Almost all turbine disc failures, even when only seldom, occur, therefore, a few minutes after the start. During this time the service personnel should not remain in the turbine disc plane, e.g., adjustment work.
With good approximation:
15°C
Temperature rise means half creep life in the region of typical operation temperatures of hot parts..
Note:
NORMAL START COMPRISES
LIFE CONSUMPTION OF
CA. 10 - 50 OPERATION HOURS
"Illustration 2.2-5": The service and maintenance factor (Lit.2-5 and Lit. 2-6) of a gas turbine is dependent, in the first place, on the load on the hot parts and, here again, at the high pressure turbine ( "Ill. 0-3"). Here the component temperature particularly determines the length of life. A rise in temperature around 15°C signifies the halving of the life under creep load (creep life, "Ill. 2.3-2"). In thermal fatigue, ( "Ill. 3.3-16"), the cyclical stress peaks are determined by the local peak temperatures. Oxidation and hot gas corrosion of the hot part surfaces, as well as micro structural damages, are clearly temperature dependent.
The number of starts is particularly noticeable at the thermal fatigue load and the cyclical centrifugal stress of the rotor components. The above diagram shows how, by the same number of running hours or rather life in relation to the number of starts, the damages on the components, respectively, the repair effort increases.
If one observes the outgoing output, i.e., the operation load, one discerns creep ( "Ill. 2.3-1"), first of all, as being a determinant to lifetime. The already named relationship, represented through an exponential mounting curve ( "Ill. 2.2-3 and 4"), stands for the maintenance effort at high peak load . The problem is relative, however, when these peaks, in comparison to the entire operation time, appear only seldom and briefly.
