Deadweight Tester

What is a Deadweight Tester

A deadweight tester is a mechanical instrument that produces a known reference pressure by placing calibrated weights on a vertical piston that floats inside a close-fitting cylinder filled with a pressurized fluid. The fluid pressure pushes the piston upward; the weights push it downward through gravity. When the two forces balance and the piston floats freely, the generated pressure equals the downward force divided by the effective area of the piston-cylinder assembly. This is a direct application of Pascal's law: pressure applied to a confined fluid is transmitted equally in all directions. Because the pressure is computed from first principles of mass, gravity, and area rather than read from a transducer such as a strain-gauge load cell, the deadweight tester is classified as a primary standard.

The governing relation is pressure equals force divided by area, often written P = F / A. The force is the weight of the calibrated masses plus the piston, corrected for the buoyancy of the surrounding air. The area is not the simple geometric area of the piston but the effective area, defined as the mean of the piston cross-section and the cylinder bore at the reference temperature, because the pressurized fluid acts across the small annular gap between piston and cylinder. The effective area is the single most carefully characterized property of any deadweight tester, and it is what an accredited laboratory certifies when the instrument is calibrated.

Structurally, a deadweight tester has three functional groups. First, the piston-cylinder assembly, machined to micrometre-level clearance and lapped for a smooth, low-friction fit. Second, the weight set, a stack of calibrated masses trimmed so that defined nominal pressures appear when specific weights are loaded. Third, the pressure-generation system: a screw pump or volume adjuster for hydraulic units, or a regulated gas supply for pneumatic units, plus the manifold that connects the reference piston to the device under test. During a measurement the operator spins the piston so that it rotates and floats on a hydrodynamic film, which all but eliminates static friction between piston and cylinder and lets the assembly settle to its true equilibrium height.

The industrial history of the deadweight tester runs alongside the history of precision pressure itself. After Eugene Bourdon patented the elastic-element pressure gauge in 1849, industry needed a way to verify those gauges, and the piston gauge emerged as the reference against which Bourdon tubes were checked. Through the twentieth century national metrology institutes refined piston-cylinder geometry, gravity determination, and effective-area modelling, and today institutes such as NIST in the United States and PTB in Germany maintain primary deadweight testers as the realization of the pressure unit over wide ranges. Commercial reference benches from Fluke Calibration, DH-Budenberg, Ametek, Mensor, and WIKA bring the same principle into accredited and industrial laboratories.

In application scale, deadweight testers span an enormous pressure window, from roughly 15 mbar at the low end of pneumatic vacuum and low-pressure models up to about 4000 bar (60,000 psi) on dedicated high-pressure hydraulic benches. No single instrument covers that whole window; the range is divided among pneumatic and hydraulic designs and among individual piston-cylinder modules, each optimized for a sub-range. The engineering essence of choosing and using a deadweight tester is therefore matching the piston-cylinder, the fluid, and the correction model to the pressure range and the accuracy the work demands.

Types and Classification

Deadweight testers are classified first by the pressure-transmitting fluid, which is the decision that most strongly shapes range, cleanliness, and low-pressure accuracy. The two families are pneumatic (gas-operated) and hydraulic (liquid-operated), with high-pressure hydraulic units forming a distinct third group. A secondary classification is by metrological grade, from field testers used at the gauge bench to primary standards held at national institutes. The table below compares the main families on the parameters that drive selection.

Pneumatic deadweight testers use a regulated gas, typically clean dry air or nitrogen, to support the piston or, in floating-ball designs, a precision ball nested in a nozzle. Gas operation gives better accuracy and stability than oil below about 6 bar (90 psi) and can reach down to roughly 15 mbar, which oil-based units cannot do because liquid head and surface tension dominate at very low pressure; below that floor a liquid-column manometer becomes the practical low-pressure reference. Gas also leaves no liquid residue, so pneumatic testers are preferred for calibrating low-range transmitters, gas instruments, and oxygen-service devices where oil contamination is forbidden. The practical ceiling of pneumatic testers is typically 100 to 120 bar. The Ametek PK II is a representative floating-ball pneumatic standard in which the ball and weights float on a thin, virtually frictionless film of gas maintained by a built-in flow regulator, removing the need to spin the weights.

Hydraulic deadweight testers use oil, most commonly a light spindle oil, to transmit pressure to a piston, with a screw pump or volume adjuster building the pressure until the loaded piston floats. Oil units are the workhorse of the medium and high range, reaching about 1400 bar (20,000 psi) in the Fluke P3100 oil configuration. They perform poorly below about 1 bar, where the hydrostatic head of the oil column and surface tension become significant relative to the generated pressure, so the low end belongs to pneumatic instruments.

Water-operated hydraulic testers substitute distilled water for oil where the calibrated devices or downstream process cannot tolerate hydrocarbon contamination, for example in food, pharmaceutical, or certain clean process applications. The Fluke P3200 water-operated tester covers ranges to about 700 bar (10,000 psi). Water introduces its own handling considerations, including corrosion control and the temperature dependence of water density, but removes the residue problem of oil.

High-pressure hydraulic testers are a specialized hydraulic subgroup built for the upper extreme. The Fluke P3100 HP and P3800 series reach 30,000, 40,000, and 60,000 psi (about 2000, 3000, and 4000 bar) across their P3830, P3840, and P3860 models, serving high-pressure hydraulics, oil and gas wellhead equipment, and pressure-vessel test rigs. At these pressures the elastic distortion of the piston-cylinder under load becomes a first-order effect, so the effective-area distortion coefficient is characterized and applied rather than ignored.

Piston-Cylinder Technology and Materials

The piston-cylinder assembly is the heart of a deadweight tester, and its design and material set the accuracy grade, the recalibration interval, and the price. Three construction approaches dominate: the simple piston-in-cylinder used in most field and workshop testers, the re-entrant or controlled-clearance assembly used in reference and primary standards to manage elastic distortion, and the floating-ball nozzle used in some pneumatic instruments. The table below compares the common material pairings and the engineering reason each is chosen.

Tungsten carbide piston-cylinders are the choice for metrology and reference benches because of their high hardness, low friction coefficient, small thermal expansion, and excellent resistance to wear and elastic deformation. These properties keep the effective area stable across thousands of pressurization cycles, which is precisely what lets a high-end tester hold 0.015 percent of reading and stretch the recalibration interval to several years. For a tungsten-carbide piston operating in a steel cylinder the combined thermal correction of the effective area is about 17 ppm per degree Celsius, so a 1 degree Celsius temperature error contributes only 0.000017 of the reading. An all-carbide pair lowers the area temperature coefficient further, toward 9 ppm per kelvin in published measurements, and is favoured on high-pressure reference benches where distortion must be minimized.

Hardened stainless steel assemblies are used in general-purpose hydraulic field and workshop testers. They are corrosion resistant and considerably cheaper than carbide, at the cost of a larger thermal expansion coefficient and somewhat faster wear, which is acceptable at field accuracy grades of 0.025 to 0.1 percent. Steel-in-brass pairings appear in legacy and budget designs; they are serviceable for routine gauge checks but have higher and less stable temperature behaviour, so they are not used where traceable low uncertainty is required.

The float position and fall rate are practical indicators of the assembly's health. In normal operation the piston is spun and settles to a defined float position within its working travel, and the operator reads pressure only while the piston floats in that band. A small, steady downward drift, the fall rate, is normal and reflects the controlled leakage of fluid through the annular gap; an excessive fall rate signals wear, contamination, or a damaged piston. If the temperature rises, the piston floats a little higher as the fluid expands; if there is a small leak it floats a little lower, and the float-position window is what tells the operator the measurement is valid.

Weight sets are calibrated masses, normally machined from non-magnetic stainless steel with a density near 8.0 grams per cubic centimetre so that the air-buoyancy correction is predictable and well defined. Small trim or fractional weights are sometimes aluminium. The masses are trimmed at the factory either to read true pressure at a stated standard gravity, or as true mass values, depending on the calibration philosophy of the maker. A weight set should never be mixed between instruments, because the mass values are matched to a specific piston effective area; swapping weights invalidates the certified relationship between load and pressure.

Corrections and Traceability Standards

A deadweight tester reads true only after a defined set of corrections is applied, and skipping them is the most common cause of an out-of-tolerance certificate. The correction model is not optional fine print: at the 0.015 percent level, several of these effects each exceed the entire allowed uncertainty if ignored. The five corrections that always matter are local gravity, air buoyancy, temperature, head height, and effective-area distortion with pressure. The table below summarizes their typical magnitude.

Local gravity is the largest environmental influence and usually the dominant uncertainty contributor. Standard gravity is defined as 9.80665 metres per second squared at 45 degrees latitude and sea level, but the true local value changes with latitude and altitude. Phoenix, Arizona, for instance, measures about 9.79474 metres per second squared, a difference of more than 1200 ppm or 0.12 percent, which dwarfs the 150 ppm uncertainty of a typical instrument. The fix is to apply a gravity correction equal to local g divided by the calibration g stated on the certificate, or to order weights trimmed for the site gravity. Local g can be obtained from a national survey or measured on site.

Air buoyancy reduces the effective load because the surrounding air exerts an upward force on the weights, equal in fraction to roughly the air density divided by the weight density. For stainless steel weights in normal laboratory air this is about 150 ppm, so it must be applied in gauge mode. The exception is absolute-pressure operation with a vacuum bell jar over the weights, evacuated by a vacuum pump and monitored with a vacuum gauge, where there is no air to buoy them and the correction is omitted. Because air density depends on barometric pressure, temperature, and humidity, those three quantities are recorded during a careful calibration, as the DKD-R 6-1 guideline requires for piston-gauge work.

Temperature corrects the growth of the effective area as the piston and cylinder expand. For a tungsten-carbide piston in a steel cylinder the combined area coefficient is about 17 ppm per degree Celsius; published all-carbide assemblies measure closer to 9 ppm per kelvin, while stainless steel pairs run higher, around 20 to 25 ppm per kelvin. The piston-cylinder temperature is read at the assembly and the area is corrected to the reference temperature. Head height corrects the hydrostatic pressure difference between the tester reference level and the reference level of the device under test; the two reference points should ideally be set to the same height when both pistons float, and modern terminals compute the residual head correction automatically.

Effective-area distortion accounts for the elastic deformation of the piston-cylinder under load, which slightly enlarges the area as pressure rises. The EURAMET cg-3 calibration guide for pressure balances recommends measuring the effective area at about ten pressures and using a linear least-squares fit to determine the area at zero pressure and the pressure-distortion coefficient. EURAMET cg-3 describes two methods, one that determines the pressure generated by an assembly directly and one that determines the masses and the effective area separately, and it specifies the cross-floating method used when the reference standard is itself a pressure balance. Alongside cg-3, the German DKD-R 6-1 guideline defines the cycles, hold times, and point counts for calibrating pressure gauges with such a standard, and OIML recommendations govern the mass and density of the weights. Traceability then flows from a national institute primary standard down through accredited working standards to the field instrument, with the certified quantities being the effective area and the individual weight masses.

Key Specification Parameters

Reading a deadweight tester datasheet is different from reading a sensor datasheet, because most of the headline numbers describe a mechanical assembly and a correction model rather than an electronic signal chain. Seven parameters truly drive the selection decision: accuracy expressed as percent of reading, pressure range and span coverage, operating fluid, piston-cylinder material, resolution and smallest pressure increment, the float and fall-rate behaviour, and the certified correction coefficients. Each is explained below.

Accuracy on a deadweight tester is almost always quoted as percent of reading, not percent of full scale, and this distinction is decisive. Percent of reading means the relative error is constant across the span, so a 0.015 percent instrument is just as accurate at 10 percent of range as at full range, whereas a percent-of-full-scale sensor loses relative accuracy as the reading falls. Field testers are typically rated 0.05 to 0.1 percent of reading, mid-range workshop standards reach 0.025 percent, and metrology benches reach 0.015 percent. With software correction that accounts for gravity, temperature, head, and air density, premium instruments such as DH-Budenberg units reach 0.008 percent of reading, with 0.006 percent offered as an option on selected piston-cylinders, and the Fluke and Ametek reference lines quote 0.015 percent of reading with NIST traceability.

Pressure range and span determine how many piston-cylinder modules and weight sets you need. A single piston-cylinder covers only a portion of the full span, so wide-range coverage is built up from interchangeable modules: pneumatic units cover roughly 15 mbar to 120 bar, oil hydraulics reach about 1400 bar, water hydraulics about 700 bar, and high-pressure hydraulics up to 4000 bar. Order the module set so that your normal calibration points sit comfortably within each module's working band rather than at the extreme low end, where buoyancy and head dominate.

Operating fluid is both a performance and a compatibility specification. Gas gives the best low-pressure performance and leaves no residue; oil gives the highest pressures; water avoids hydrocarbon contamination. Fluid choice also feeds the correction model, since fluid density sets the head correction and, in the case of water, varies measurably with temperature. Piston-cylinder material, covered in Chapter 3, sets the accuracy grade and the thermal coefficient and should appear explicitly on the datasheet for any reference-grade purchase.

Resolution and the smallest pressure increment are governed by the smallest trim weight in the set, because pressure is changed in discrete steps of mass rather than continuously. A fine increment lets you place a calibration point closer to a nominal value; a coarse increment forces interpolation and adds uncertainty. Float and fall-rate behaviour, also from Chapter 3, defines the valid measurement window: the operator reads only while the spun piston floats within its band, and an abnormal fall rate is the first sign that the assembly needs service. Finally, the certified correction coefficients, namely the effective area at reference conditions, the temperature coefficient, and the pressure-distortion coefficient, are not marketing numbers but the working constants you enter into the calculation; a reference-grade tester is only as traceable as the calibration certificate that supplies them.

One parameter that does not appear on the datasheet but governs everyday use is the recalibration interval. Because a deadweight tester is a primary standard with no aging electronics, a properly handled instrument can hold its effective area for several years, and intervals of 1 to 5 years are common, with metrology benches typically on a 2 to 5 year cycle. This long interval, combined with percent-of-reading accuracy, is why a deadweight tester remains the reference even though digital calibrators, including the handheld multifunction process calibrator, are faster to operate in the field.

Selection Decision Factors

To turn the preceding chapters into a specific purchase, work through the decision sequence below. Most selection mistakes come not from one wrong answer but from deciding the pressure range and fluid before understanding the accuracy grade the work requires. These eight steps can serve as a fixed RFQ template for a primary pressure standard.

1. Pressure range and span coverage: List the lowest and highest pressures you must calibrate and the points in between. Map them onto piston-cylinder modules so each normal point sits inside a module's working band, and decide how many modules and weight sets are needed to cover the whole span.
2. Fluid family: Choose pneumatic for low pressure (below about 6 bar), gas instruments, and oxygen-clean work down to roughly 15 mbar; choose oil hydraulic for the medium and high range to about 1400 bar; choose water hydraulic where hydrocarbon contamination is forbidden; choose high-pressure hydraulic for work above 2000 bar.
3. Accuracy grade: Distinguish field checks (0.05 to 0.1 percent of reading), workshop standards (0.025 percent), and reference or laboratory standards (0.015 percent or better, down to 0.008 percent with software correction). Each grade step adds cost and demands a more complete correction model.
4. Piston-cylinder material: Specify tungsten carbide for reference and high-pressure benches to minimize wear and thermal and elastic distortion; hardened stainless steel is adequate for general field and workshop use. Require the temperature and distortion coefficients on the certificate.
5. Correction and software support: Confirm the instrument is supplied with, or can compute, local-gravity, air-buoyancy, temperature, head, and area-distortion corrections, ideally through a terminal or software so the operator does not calculate head manually. For accredited work, verify the certificate states the calibration gravity, temperature, and air density.
6. Site gravity and environment: Obtain your local gravity value and decide whether to order weights trimmed for site gravity or apply a software correction. Record laboratory temperature, barometric pressure, and humidity for the buoyancy correction, as DKD-R 6-1 requires.
7. Traceability and standards conformance: Require NIST, PTB, or other national-institute-traceable calibration, conducted per EURAMET cg-3 for the pressure balance and DKD-R 6-1 for downstream gauge calibration, with the effective area and weight masses certified and an uncertainty statement attached.
8. Total cost of ownership: Add the piston-cylinder modules, weight sets, pump or gas supply, software, and the multi-year recalibration cost. A deadweight tester's long recalibration interval and stable effective area usually give a lower lifetime cost than a sensor-based calibrator that must be recertified every 1 to 2 years, despite the higher purchase price.

One last dimension that buyers often overlook is serviceability and the supplier's metrology chain: whether the maker operates accredited laboratories that send master testers to a national institute on a regular interval, whether spare piston-cylinders and weights are available, and how field recalibration is handled. Ametek states that its testers are directly NIST-traceable, with master testers sent to NIST regularly and working masters used to calibrate every new and repaired unit, while Fluke Calibration, DH-Budenberg, Mensor, and WIKA maintain comparable laboratory chains. These factors seem secondary at purchase but determine how easily the standard stays traceable across a 10 to 20 year service life.