Harwood Engineering Company, Inc.

Manganin High-Pressure Sensors

by Donald H. Newhall

Manganin wire provides a linear increase in resistance with pressure, which is used in measuring pressures up to 400,000 psi.

Elastic-deformation types (Bourdon gage, helical gages, strain gages, and bulk-modulus cell) depend on Hooke's Law. But none follows Hooke's Law precisely, and all have hysteresis of 0.1% or more.

The manganin cell depends on change in the resistance of a coil of wire when the coil is exposed to a high pressure. The change in resistance of the coil is measured by a bridge.

Manganin alloy has a large, positive, linear pressure/resistance relationship* and, under ordinary conditions, is relatively insensitive to temperature fluctuations. Manganin cells can yield accurate measurements to over 400,000 psi, but are delicate (other materials, such as gold-chromium or lead, also have been used).

The manganin pressure gage has long been used in high-pressure work. Bridgman² describes it and relates the circumstances leading to his adoption of it as a secondary gage for general use in the broad program of research which earned for him a Nobel prize.

Following the results of Lisell³, Bridgman measured the resistivity of some well-seasoned German manganin to 13,000 kg/cm² using a free-piston gage. He found the manganin to be linear to within 0.1%; its usefulness for a pressure gage was immediately apparent. Placing a coil of the wire within a press, Bridgman incorporated it in a Carey-Foster bridge and converted directly into measures of pressure the bridge unbalance (multiplied by a predetermined constant). He later extended his pressure scale to 30,000 kg/cm² again using a free-piston gage as a standard.⁴ The same linearity was found at this pressure. Other experimenters have investigated the resistivity of manganin at high pressures, with comparable results.⁵

The pressure/temperature/resistivity relations of several alloys, and different melts of the same alloy, were studied by the author and Mr. Horace Darling of the Foxboro Company⁶. The data showed that an alloy of gold with 2.1% chromium is a promising material for a pressure gage because of its unusually low temperature effect and its linearity with pressure. However, its pressure coefficient of resistance is less than one-half that of manganin. Gages have been made of the gold-chromium alloy, but manufacturing problems have precluded its general use for an industrial pressure gage. Attaching leads to a coil of gold-chrome is difficult. A soldered or brazed joint of gold-chrome to copper, or other lead material, appears to be progressively influenced by pressure. An effect suggestive of amalgamation occurs at the interface between the gold-chrome and the solder, resulting in a tendency to brittleness, leading to rupture with the application of pressure.

The effect of pressure on the resistivity of manganin may be influenced by various factors. The pressure coefficient can vary by as much as 10%, depending on the composition of the sample.

It is most important that the wire constituting a pressure gage be free from internal strain. Such strain may be expected when the wire is wound into a coil, and the coil must be treated to remove the strain, as by annealing it at as high a temperature as its insulation will stand, and by subjecting it repeatedly to a pressure equal to that which it will encounter in use — preferably as much as 10,000 atmospheres. Unless this is don, linearity cannot be guaranteed, and the resistance of a coil will generally change after applying pressure. In extreme cases, a shift may occur for several cycles of pressure.

The pressure coefficient of resistivity of manganin depends to some extent on the temperature. Manganin is one of the alloys least sensitive to temperature, but for precise measurement the temperature should be considered.

Figure 2 shows a typical temperature-resistivity curve for manganin wire. While the shape is characteristic, that temperature at which the wire is least sensitive to temperature change (the point of zero slope) depends on the amount of residual strain in the wire, as well as on its composition. After adequate seasoning, this point should be at room temperature. Within about five degrees the coil should be insensitive to temperature change except for measurements of the highest precision.

The pressure-resistance sensitivity of manganin is essentially constant for any one melt. Manufacturers have been known to wind wire from two melts on the same spool; if the user encounters a splice he should make a determination of the pressure coefficient of the wire beyond the splice.

The manganin high pressure cell (Figure 1) has two noninductively wound resistance coils. One is active and the other is a compensating coil. The active coil is subjected to the hydrostatic pressure to be measured, and the other is located in the cap of the cell. A single electrically-insulated lead s brought into the cell through the cell plug. The body of the plug connects to the other side of the active coil. Thus the plug is much like a spark plug in appearance.

The two coils are wired in series to form adjacent arms of a Wheatstone bridge. The wire currently in use at Harwood has a pressure sensitivity of about 1.65 x 10^-7 ohms/ohm/psi. Both coils have 120 ohms (± 0.1 ohm) resistance at atmospheric pressure. Thus, for 100,000 psi there will be a resistance increase of:

The cell is used commonly with a Dynalog (Foxboro) recorder with an a-c bridge, or with a Harwood Carey-Foster d-c bridge.

The Foxboro instrument has three ranges, such as 0 to 50,000, 0 to 100,000, and 0 to 200,000 psi. There is a range selector switch, an adjustable gage factor and a zero adjustment in the instrument. It is accurate and readable to approximately 0.5% of scale.

When greater accuracy is desired, and when small changes in pressure must be observed, the d-c Harwood Carey-Foster bridge is more suitable. Pressure changes of 15 to 25 psi in 200,000 psi may be observed with this equipment. The manganin gage is linear in its resistance change in response to pressure to better than 0.1%.

The compensating coil compensates for gross ambient temperature changes. Adiabatic changes within the cell will affect the pressure reading slightly. A five-minute dwell period at pressure with an electronic instrument, or a twenty-minute period with the more-sensitive and precise d-c technique may be required to obtain temperature equilibrium.

Manganin pressure cells are available in two models. The larger unit is designed so that the fluid in the high-pressure system is isolated from the sensing coil. This is necessary when the fluid is conductive, corrosive, or could become conductive; for instance, if it were hygroscopic. The isolation is accomplished by a bronze bellows around the coil. The bellows is vacuum filled with a suitable non-conducting fluid that will not solidify under pressure.

The smaller unit is built into a tee or cross of a 12H 200,000-psi fitting, and is suitable for experimental work in which care is taken not to allow fluids in the pressure system that would harm the coil.
 

References