gsphelp  Clutch

<< Click to Display Table of Contents >>

Navigation:  GSP components > Component Libraries > Miscellaneous Components > STOVL Component Library >

gsphelp  Clutch

Previous pageReturn to chapter overviewNext page

STOVLClutch10

 

The clutch component is used to transmit the power from the main drive shaft to the lift fan drive shaft in STOVL operation. Engaging and disengaging the lift-fan driveshaft from the fan shaft while the engine is running can be modelled.

 

A clutch is an unusual element in a gas turbine drive train system. However, there are a number of applications, where the load needs to be coupled and uncoupled from a gas turbine drive shaft during operation. In these cases, a clutch is required to allow smooth transfer from the uncoupled state to the fully coupled state and vice versa, without excessive torque loads on the shafts due to high acceleration rates.

Examples of clutches in gas turbine drive trains may be found in: vehicular turboshaft engines, helicopter drive trains and STOVL propulsion systems using liftfans driven by the main engine LP shaft.

 

In general, a clutch is a device that is able to transfer a certain (limited) amount of torque between two shafts. Several systems exist to transfer the torque including wet and dry friction plate systems and a variety of hydraulic systems (?).

 

For analysis of system performance, the clutch model minimally needs to accurately represent the torque transferred. If also clutch performance itself requires scrutiny, more detailed models may well be required. This section describes the clutch model as implemented in GSP (NLR Gas turbine Simulation Program, ref. xx), with calculation of torque transmission, clutch state and friction heat production.

 

For the clutch model, a number of terms/parameters are introduced required to determine the state of operation of the clutch.

1.Engagement status
A clutch can be fully engaged of disengaged. If fully engaged, it is able to transfer maximum torque capacity; if fully disengaged, usually no torque is transferred unless some sort of residual friction loss is defined in the model.
In GSP an 'engagement variable' is used, ranging from 0 to 1. 0 is fully disengaged, 1 means fully engaged.
2.Locked/unlocked status
If a clutch is locked, both shafts run at the same rotor speeds and the clutch functions as a coupling. In this case, the torque transferred does not exceed maximum static torque capacity. If the clutch is unlocked, both shafts are not running equal speeds. There is a case where the clutch in unlocked at equal speeds, but this only can occur during a very short time of transition between the locked and unlocked states, or when at least one shaft is accelerating past the other shaft speed.
3.Static torque capacity
Static torque capacity is the maximum torque the clutch can transmit in the locked state. This means the friction materials do not move (relative to each other) and the static friction coefficient applies. Static torque capacity always is equal or larger than dynamic torque capacity.
4.Dynamic torque capacity
Dynamic torque capacity is the maximum torque the clutch can transmit when it is unlocked, i.e. rotor speeds are not equal and so the friction materials are moving relative to each other, so the dynamic friction coefficient applies. Dynamic torque capacity always is equal or smaller than static torque capacity.
5.Torque demand
Torque demand is the torque that would be transmitted in the locked state. It can also be described as the torque that would exist in the shaft if maximum torque would be infinite and no slipping would occur.
6.Slipping
The clutch is slipping if the two rotor speeds are unequal and engagement is larger than 0. Torque required exceeds maximum dynamic torque capacity and therefore cannot fully be transmitted. This means friction heat is produced proportional to the torque (engagement x maximum dynamic torque) and the delta in rotor speeds.

 

The operation mode of interest with a clutch model is transient. For steady-state simulation, either the fully engaged or disengaged state must be assumed. For system modeling environments like GSP, this means that the design point state also either is fully engaged or disengaged. Prior to a transient simulation, then a fully engaged or disengaged state must exist.

 

GSP implementation

In the GSP clutch component, a static and dynamic maximum torque can be specified. The engagement value, ranging from 0 at disengagement to 1 at full engagement, determines the actual torque capacity as a fraction of maximum torque.

The engagement value during a transient simulation can either be obtained from user-specified time functions or result from a control system model output. With a clutch control model, accurate simulation of clutch performance in complex systems including (closed loop) engagement can be performed.

Friction heat production due to clutch slipping is calculated, and with more data an accurate heat flow and conduct model can be added to analyze local heat loads and temperature levels during and after successive engagement events

In ref. GSP and example is given of simulation of a turbofan driving a STOVL lift-fan through a clutch. The lift-fan is driven by the main engine fan shaft through a dry Clutch that is able to disengage the Lift Fan from the engine during normal forward flight and engage during vertical flight modes. The GSP model and libraries required are included in the GSP registered version.

 

One of the challenges in modeling clutch engagement- or disengagement transients, is to determine whether static or dynamic maximum torque is applicable. In other words: is the clutch in a locked state or is it slipping. In the transient simulation this means the model algorithm needs to keep track of the clutch state (locked or not) history during transient simulation, i.e. save the clutch state of the prior time step. The state of the prior time step then determines whether dynamic or static maximum torque applies. For example, during a transient starting with a locked clutch but increasing torque demand, at some point this torque will exceed maximum static torque and the clutch will unlock and start slipping. After the time step where the unlocking has been determined, the lower level of dynamic maximum torque applies, and the clutch can only lock again after the torque demand will drop below dynamic maximum torque. If torque demand fluctuates around maximum static torque levels, severe transient load effects on the clutch can be simulated. In this case however, small time steps will be required to accurately represent rapid transient effects.