Control valve mechanical test procedure

I was asked to create this control valve mechanical test procedure for one of the world’s leading engineering companies.

1 Introduction
2 Location of tests in the manufacturing programme
3 Stem position error test
4 Deadband test
5 Hysteresis test
6 Hysteresis plus deadband test
7 Stroking time test
8 Operation instructions
9 Equipment specifications

1 Introduction

The company has the facility to provide a permanent record of an assembled valve’s performance, whilst undergoing mechanical operation tests. The results of these tests are recorded on A3 size paper using an XYT pen recorder.

2 Location of tests in the manufacturing programme
All tests are carried out on the assembled valve and actuator before hydrostatic and seat leakage testing. The packing box gland flange nuts are finger-tight and careful assembly of the valve ensures that the packing material is in an uncompressed state. A light lubricating oil is applied to the area of the stem that passes through the packing box.

3 Stem position error test
The purpose of the stem position error test is to verify that the desired and actual stem positions are within acceptable limits. The acceptable stem position error for the company’s range of control valves is ± 5% of the rated travel. The stem position should lie between the following upper and lower limits for an input signal of 3 to 15 psig:

INSTRUMENT	% OF RATED TRAVEL
AIR PRESSURE	LOWER	UPPER
(PSIG)	LIMIT	LIMIT
3		5.00
4	3.33	13.33
5	11.67	21.67
6	20.00	30.00
7	28.33	38.33
8	36.67	46.67
9	45.00	55.00
10	53.33	63.33
11	61.67	71.67
12	70.00	80.00
13	78.33	88.33
14	86.67	96.67
15	95.00

The equipment is connected to the valve and actuator as described in section 8.

4 Dead band test
Deadband is the range through which an input can be varied without initiating observable response. In a diaphragm actuated control valve, deadband is the amount that the instrument air signal can be changed without initiating valve stem movement.
The amount of deadband is determined by measuring the changeover pressure for a given stem position. The stem is taken up to a position of 25% of the rated travel. While stationary, the change in pressure which causes a change in stem movement is measured. The test is repeated at positions of 50% and 75%. This changeover pressure is known as the 'deadband' and it should not exceed the following values :

VALVE AND DIAPHRAGM ACTUATOR WITHOUT POSITIONER
SPRING RANGE	POINT	MAXIMUM CHANGEOVER
	OF TEST	PRESSURE
3 TO 15 PSIG	25%	0.20 PSI
	50%	0.25 PSI
	75%	0.35 PSI
6 TO 30 PSIG	25%	0.40 PSI
	50%	0.50 PSI
	75%	0.70 PSI

VALVE AND DIAPHRAGM ACTUATOR WITH POSITIONER
SPRING RANGE	POINT	MAXIMUM CHANGEOVER
	OF TEST	PRESSURE
3 TO 15 PSIG	25%	0.0056 PSI
	50%	0.0068 PSI
	75%	0.0096 PSI
6 TO 30 PSIG	25%	0.0112 PSI
	50%	0.0140 PSI
	75%	0.0192 PSI

The equipment is connected to the valve and actuator as described in section 8.

5 Hysteresis test
Hysteresis is a characteristic of a control valve that is the dependence of the stem position, for a given variation of the instrument signal, upon the history of previous variations and the direction of the varying instrument signal, i.e. increasing/decreasing. The amount of hysteresis is determined firstly by performing the deadband test, followed by stroking the control valve over its full travel and returning it to its starting point. The amount of hysteresis is calculated by deducting the deadband from the distance between the cyclic envelope at 25%, 50% and 75% travel. The hysteresis should not exceed the following :

POINT	MAXIMUM
OF TEST	HYSTERESIS
25%	0.40 PSI
50%	0.35 PSI
75%	0.25 PSI

The equipment is connected to the valve and actuator as described in section 8.

6 Hysteresis plus deadband test
Hysteresis plus deadband is the total dynamic friction present in a control valve and is the vertical or horizontal distance between the cyclic envelope obtained from the hysteresis test.  The acceptable hysteresis plus deadband for a control valve is ±5% of the rated travel.

The equipment is connected to the valve and actuator as described in section 8.0.











7 Stroking time testThe stroking time of a valve is the time taken for the valve to stroke over its entire travel. This may be from the fully open position to the fully closed position, or vice versa. The duration is measured from signal increase/ decrease to full travel.  As the stroking speed is dependant upon many factors, it is not practical to define acceptable limits. The influencing factors listed in order of priority are :

(a) Actuator size
(b) Actuator stroke
(c) Air supply
(d) Pressure
(e) Size of pipework connections
(f) Spring rate
(g) Air to open/close
(h) Type of positioner

The equipment is connected to the valve and actuator as described in section 8.

8 Operation of the test equipment
The plugs and sockets of the electrical equipment are individually numbered for ease of assembly and to eliminate the possibility of incorrect wiring. Plugs and sockets, having the same number, should be connected together - with careful consideration of the actuator fail position. There are two leads labelled 'R', and a further two labelled 'D'. When testing equipment having a reverse acting actuator, the leads labelled 'R' should be connected and when testing equipment having a direct acting actuator, the leads labelled 'D' should be connected.

The power supply for the pressure transducer is situated on the left of the cabinet and it is labelled 'PRESSURE TRANSDUCER'. The operating voltage of this power supply should be set at 10 VDC and this is achieved by careful adjustment of the coarse and fine potentiometers. In most cases, the pressure transducer power supply will already be set at exactly 10 VDC.

The power supply for the linear potentiometer is situated to the right of the pressure transducer power supply and it is labelled 'LINEAR POTENTIOMETER'. The operating voltage of this power supply should be set at 24 VDC and this is achieved by careful adjustment of the coarse and fine potentiometers. In most cases, the linear potentiometer power supply will already be set at exactly 24 volts VDC.

The scale setting for the instrument signal axis is situated on the left hand side of the pen recorder and it is labelled 'INSTRUMENT SIGNAL'. Three controls are required to be set and their positions are dependant upon the maximum instrument signal pressure used.

For a maximum instrument signal pressure of 15 psig, set the range knob to 2mV/cm, vernier to (approximately) 0.00 - 0.40, zero adjustment to (approximately) 4.90 - 5.10. The vernier and zero settings are approximate and may require fine adjustments to achieve full scale deflection.

For a maximum instrument signal pressure of 30 psig, set the range knob to 2 mV/cm, vernier to (approximately) 7.08, zero adjustment to (approximately) 4.19. The vernier and zero settings are approximate and may require fine adjustments to achieve full scale deflection.

The scale setting for the valve stroke axis is situated on the right hand side of the pen recorder and is labelled 'VALVE STROKE'. Three controls require setting and their positions are dependant upon the maximum valve travel.

For a maximum valve travel of 1.1/8", set the range knob to 0.1 V/cm, vernier to(approximately) 6.72, zero adjustment to (approximately) 5.20.
For a maximum valve travel of 1.1/2", set the range knob to 0.1 V/cm, vernier to (approximately) 7.63, zero adjustment to (approximately) 5.20.
For a maximum valve travel of 2.1/4", set the range knob to 0.2 V/cm, vernier to (approximately) 6.72, zero adjustment to (approximately) 5.20.
For a maximum valve travel of 3.1/2", set the range knob to 0.5 V/cm, vernier to (approximately) 1.76, zero adjustment to (approximately) 5.11.

All of the above settings for the vernier and zero adjustments are approximate only and may require fine adjustments to achieve full scale deflection.

9 Equipment specifications
Linearity of the linear potentiometer is better than 1%, as is the linearity of the pressure transducer.

EQUIPMENT	MODEL NUMBER	SERIAL NUMBER
Pen recorder - Farnell	RW101	F2058
Power supply (Potentiometer)	E30/1	005949
Power supply (Transducer)	E30/1	006189
Pressure transducer - Honeywell	136PC30G1
Linear potentiometer - Penny & Giles	LP26/200/6"/6K	108112B

Case Study : Magazine System for Robot Assembly

This is a  design proposal that I was asked to carry out for a Swedish world leading manufacturer of compressors, generators, construction and mining equipment, industrial tools and assembly systems.  They required a magazine system to present end-pieces for the pre-assembly of pneumatic cylinders.

The second stage for the robot assembly of pneumatic cylinders involves the sub-assembly of end-pieces and half-pistons.  End-pieces need to be handled by a magazine system because they are too large for conventional vibratory feeders.  A magazine system is required at the pre-production facilities that is a scaled-down version of the future production system, within budget limitations.  The cost of the system is split between the fixed cost for the transfer of parts to the robot and the variable cost of end-piece storage.  The variable storage cost is proportional to the capacity of the magazine system. There is also an indirect labour cost for the filling and transport of magazines, in addition to the equipment material cost. The prototype can have the same transfer device as the production model, but with a smaller capacity.

MAGAZINE FILLING

The only economical method of magazine loading is to fill them at the point of final manufacture. This is because the time taken to insert a part into a magazine can approach the time taken to insert it into the part-built assembly.  Nevertheless, end-pieces have to be transported from manufacture to assembly and magazines are the best way of doing this, whilst also giving protection to the surface finish.

MAGAZINE CAPACITY

The capacity of the magazine is as large as possible to achieve the minimum number of journeys from manufacturing to assembly during the shift.  If demand for each cylinder diameter is equal then the magazine must contain in excess of sixty parts for a refill only once a shift.  A single vertical stack magazine would be in excess of three metres high. It is therefore proposed that a number of units should be combined to form one magazine. Three magazines of twenty end-pieces seems reasonable.

PROPOSED SYSTEM

The system shown is one method of end-piece distribution. The illustration shows one magazine to store one style of end-piece.  The production version for the Swedish manufacturing plant would have three magazines per end-piece, each behind the another.