New torsion testing e quipment designs offer state-of-the-art instrumentation and provide enhanced capabilities including greater speed, measurement accuracy, and flexibility of experiment design.

Metals workability generally is defined as the ability to form a metal into a particular shape by means of applying a load to the metal via some type of equipment or tooling without causing the metal to fracture or to have an undesirable microstructure. Workability of a material is determined from several factors including strain, strain rate, temperature, failure behavior, and others. Mechanical tests, such as tension and compression tests alone are not capable of providing all of this information. However, the torsion test can provide flow stress data and failure and microstructural response to deformation processing. Gleeble systems offer dynamic thermal-mechanical test capabilities, such as hot torsion testing, simulating thermal and mechanical parameters under very precisely controlled conditions.

Fig. 1 Results of a three-hit deformation simulation.

Hot Torsion Test

The Gleeble Hot Torsion System has a maximum rotation speed of 1500 rpm and a maximum torque cell rating of 56 Ncm (500 lbf in), typically large enough for hot torsion testing of 10-mm (0.4-in.) diameter steel specimens. A small torque cell is used to obtain more accurate torque measurements, even though the hydraulic torque motor is rated to 100 Ncm (880 lbf in). The torque cell is overload-protected and the mounting design protects the cell from excess torque. In this way, the small load cell can be used without danger of damage from the much larger torque motor. The larger torque motor provides rapid acceleration to simulate multihit rolling processes (typically 15 msec from 0 to 1000 rpm with no load). Figure 1 shows the results of a three-hit deformation simulation for a 10 mm diameter, 50 mm long gage length (0.4 by 0.2 in.) AISI Type 1018 carbon steel specimen.

With the advantage of the large strain achieved in torsion testing, a multiple hit process simulation is readily conducted on the torsion unit as shown in Figure 1, as long as the specimen is ductile enough to complete the torsion test. With Gleeble torsion systems, one end of the specimen can be locked or free of restraint in the axial direction during torsion. The other end of the specimen is connected to the torsion motor and there is a coupler that allows the specimen to be directly coupled to the torsion motor or rotate freely up to 20 degrees. The free coupler can be used to accelerate the torsion motor rotation more quickly and with less initial strain error at the beginning of deformation. It also makes it easy to unload the torque on the specimen (creating zero residual torque) between two hits.

Fig. 2 A pure torsion test specimen (top) and a standard undeformed test specimen (bottom) for reference.

When one end of the specimen stays free in its axial direction, the torsion test is in pure shear state, and the gage length shrinks while the diameter of the specimen expands within the gage length as shown in Figure 2. The top specimen in the figure is after 20 revolutions under pure torsion. The bottom specimen is a standard 10-mm diameter torsion testing specimen having a 50 mm (2 in.) gage length. Changes in diameter and gage length affect torque and strain rate, and, therefore, the torque and torsion values must be corrected to calculate the stress and strain value.

Torsion with Tension Testing

The free end of the specimen can be locked to keep the gage length unchanged during torsion, resulting in the generation of a tensile stress in the torsion specimen in addition to shear stresses. This is accomplished via stroke control at zero position using a 4-mm (0.16 in.) coupler with the coupler in tension before testing.

It also is possible to apply a specific, controlled tensile force to the specimen during the torsion test. To apply a tensile stress onto the specimen during torsion, the thrust load cell is used as the feedback control and the mechanical mode of the main ram must be mmForce. The maximum thrust load cell rating depends on system load cell capabilities; standard thrust load cell configuration is 900 kg (2000 lb).

A constant tensile load also can be applied using an air ram. However, this method does not provide Force feedback and the thrust load cannot easily be changed once the test starts. The air ram can be manually or automatically turned on or off during the test, but a certain amount of time is required to build the air pressure. Thus, the preferred method to apply tension force is by means of the main hydraulic ram.

Fig. 3 A necked test specimen after combined tension and torsion testing (top), and a standard undeformed test specimen (bottom) for reference.

A specimen does not require a very large tensile load to initiate necking during torsion when combining tension and torsion at a high temperature. This is attributed to the combined shear and tensile stress states. Figure 3 shows a necked specimen during torsion under a constant axial tensile load of 1000 N (225 lbf).

As soon as necking occurs, torsion is localized within the necking area and the torque drops. This makes the torsion deformation nonuniform within the gage length and creates difficulties calculating the equivalent stress and strain, which is similar to tensile testing.

Torsion with Compression Testing

Combining a torsion test with compression is accomplished using different control modes as described above; i.e., stroke, thrust (or force) and air ram. With stroke control, the specimen can be compressed to a given amount, although there are physical limits to the amount of compression possible, which is based on specimen size. With a standard torsion specimen (see Figure 3), the maximum compression available is around 10 mm (0.4 in.). Other specimen sizes can be used if a larger compression is required. If it is necessary to apply a specific load, the main ram system can be programmed using force (thrust) control. The air ram system also can be used for compression during torsion.

Fig. 4 Test specimen bulging from testing under torsion and compression (top), and a standard undeformed test specimen (bottom) for reference.

One problem associated with compression during torsion is that the specimen may bulge. Bulging results in nonuniformity of torsion within the gage length and makes it difficult to calculate the torsion strain and stress. Figure 4 shows a bulged specimen after torsion and compression; the specimen deformed was a standard torsion specimen that did not have any end heating by an element furnace. In this case, the thermal gradient within the gage length may significantly contribute to the bulging. Thus, when a torsion test is performed under compression, it may be necessary to minimize the thermal gradient using proper specimen design and the end-heating element to reduce the bulge.

Fig. 5 Torsion test specimen failure under compression (top), and a standard undeformed test specimen (bottom) for reference.

A torsion specimen also can fail under compression and shear stress states. Figure 5 shows an example of a specimen failure under a compressive load of 1000 N at 800C (225 lbf at 1470F).

Fig. 6 Torque values of AISI Type 1018 carbon steel specimens tested at four different temperatures.

Figure 6 shows the results of tests conducted under a constant axial load of approximately 1000 N using the thrust load cell control. The AISI 1018 carbon steel specimen material has the lowest ductility at 750C (1380F), failing after only 3.5 revolutions at 30 rpm. The material also failed after 5 revolutions at 800C (1470F) at 30 rpm, but did not fracture at 900C (1650F) and around 1050 to 1100C (1920 to 2010F). Maximum torque does not exceed 25 Ncm at 750C (220 lbf in at 1380F) at 30 rpm although a higher strain rate will increase the torque value.

Details of the testing conditions for the above four tests are listed in Table 1.

Torsion-Specimen Heating

Parallel Resistance Heating

Parallel resistance heating is accomplished by passing electric current through two parallel paths between two jaws; it often is helpful to have a small sized (cross-sectional area) specimen. This type of heating also can be used to heat a specimen having very high resistivity, such as certain ceramic/ metallic materials, etc.

A specimen can be heated to a very high temperature in a fraction of a second. Thus, good temperature control of the elements in both paths becomes essential for the application of the technique.

For parallel heating, a correlation of the temperatures of the elements in the parallel paths can be derived using the following assumptions:
1) Both metal elements (specimen and/or end heating element) are heated uniformly and have the same heat transfer rate for heat loss
2)The cross section of the two elements within the free span are uniform


Temperature differences between the two elements and room temperature denoted as ÆT1 and ÆT2. The temperature can be controlled by either of the thermocouples on the two elements. After one temperature is controlled, the other is followed by the equation where K, p, cp are the specimen's electrical resistivity, density, and specific heat, respectively; and L2 and L1 are the lengths of the element in path 2 and the specimen in path 1 between two jaws under parallel heating.

Torsion Specimen Heating

Most torsion systems use either induction heating or furnace heating. Testing productivity is low due to the time required to heat and cool the furnace. Induction heating also presents unique problems related to repeatability based on coil position and creation of localized shear bands, which are located within the specimen based on the location of the induction coil. In an effort to minimize these issues, resistance heating is used in the Gleeble hot torsion test equipment. A much higher heating rate and more uniform cross-sectional heating can be achieved using the bulk resistance-heating method.

Proper specimen design and the use of a specially designed end heating element are required to obtain a uniform temperature zone within the specimen gage length.

Fig. 7 End heating element for torsion testing.

A specially designed end-heating element is used to keep the torsion specimen temperature uniform within the gage length, as shown in Figure 7. This is a typical application of parallel heating. When the length of the end heating element is twice the length of the free span (to keep the elements in two paths having the same temperature in the middle of the free span during heating), the value of Kpcp of the end-heating element must then be four times smaller than the value of Kpcp of the torsion specimen. This has been verified on the Gleeble system and a reasonable agreement has been achieved using stainless steel hot jaws.

Table 2 lists the properties of some common metals at room temperature for use as design references. It appears that molybdenum is a good candidate for end-heating element of carbon steel torsion specimens. When the length of the elements is about 150 mm (6 in.) and the free span of the torsion specimen is about 80 mm (3 in.), the end temperature of the element will reach the specimen temperature to keep the temperature within the gage length uniform.

To have the same temperature during parallel heating of two metal specimens, the relationship of the Kpcp between the two specimens is shown in the equation. When the two specimens are made of the same material, specimen lengths (L1 and L2) in the free span must be the same to have the same temperature. This indicates that the temperature in each path is not affected by its cross-sectional area as long as it is a uniform bar.

Fig. 8 Torsion test specimen set-up and end heating element.

Part of the heating element used on DSI's torsion unit is made of low carbon steel sheet, which works satisfactorily. Figure 8 shows the installation of the end heating element with the torsion specimen in the torsion working chamber. Two pairs of Type K thermocouples were welded within the gage length of a torsion specimen with one in the middle and the other at the edge of the gage length. The specimen is modified from a standard torsion specimen by drilling an 8.2 mm (0.312 in.) diameter hole at each end of the specimen. The holes are drilled to the shoulder of the gage length. With the hole at each end, the cross-sectional area is close to that of the gage length section, and so is current density.

Fig. 9 Schematic of a torsion test specimen containing a hole drilled to the specimen shoulder at each end (see text for details).

As a result, the end part of the specimen having a larger outside diameter can be heated as effectively as in the gage length section. Details of the modified specimen with dimensions are shown in Figure 9.

Fig. 10 Temperature difference within the gage length of a torsion test specimen

The temperature difference within the gage length was monitored during heating, soaking and cooling. Figure 10 shows the temperature difference within the gage length of a torsion specimen during heating and cooling. After heating to 1200C (2190C), the temperature difference is less than 5C (9F), and the temperature difference diminishes after several seconds. Near the end of the 1200C specimen soak period, the temperature at the specimen shoulders drops by about 10C (18F) during breakdown of the vacuum by introducing argon gas into the chamber. This is due to the fact that the shoulder has a larger surface-to-volume ratio than that in the middle of the specimen during forced argon convection.

Remarks
The hot torsion system described here has the capability to control the thermal profile of a test specimen to an extent not previously possible including rapid heating and cooling, maintaining uniform temperatures and producing temperature distributions having controlled thermal gradients. This together with the capability of performing torsion tests in combination with tension and compression allows the study of the very complex stress state and the effects of these variables on the ductility of materials-research areas for which there is a need for data.