Project Leader: G.A. Alers
S.R. Schaps

MSEL Program Intelligent Processing of Materials	MRD Focus Technology IPM - Thermomechanical Processing
Strategic Thrust Advanced Processes	Character of Research Metals Producers

Technical Description

Most material processing for commercial applications is designed to develop specific microstructures that are known to give the final product its desired properties. Therefore, the development of process-control algorithms and nondestructive evaluation procedures should be focused on developing techniques and sensors that provide information on the microstructure instead of detecting localized inhomogenities that may or may not be flaws.

Technical Objectives

The objective of this program is to investigate measurement techniques that can be applied in a production environment and that give information on the microstructure of the material being produced. Thus, the program demands establishing quantitative relationships between measurable physical properties and the microstructure, as well as relating the microstructure to the commercially desirable properties that govern the production process. The relationships must be sufficiently well defined to allow the development of models that can predict the commercial properties from the physical property values. This is actually a classic case of inversion of an ill-posed problem. Our objective is to develop models that relate a desirable commercial property (such as hardness) to the microstructure as well as a model for relating a physical property (such as sound velocity) to the microstructure, and then to invert the relationships to yield a quantitative procedure for predicting the commercial property from a physical property measurement.

FY95 Accomplishments

During the past several years, NIST has established unique capabilities in noncontact eddy current and ultrasonic transducers that can operate in the hostile environments found in industrial metal-forming operations. In particular, the sensors can measure electrical resistivity, ultrasonic wave velocities and attenuations, and some magnetic properties of hot metal sheet as it moves through a rolling mill at high speed. Thus, the emphasis during FY95 was the development of techniques for: (1) making precision ultrasonic wave velocity measurements on thin polycrystalline sheet metal; (2) establishing relationships between the velocities and the microstructure of the individual grains; and (3) investigating correlations between ultrasonic properties and the mechanical strength properties (yield strength, hardness, grain size, etc.) desired by the consumers for the sheet product. In the case of sheet steel, magnetic property measurements can also yield information on the microstructure. Therefore, development of magnetic property measurements that are compatible with the ultrasonic techniques and with rolling mill operations are being undertaken.

Task I. Simulation of On-line Testing. In order to demonstrate that the ultrasonic, magnetic and electrical properties of thin sheet can be measured under rolling mill conditions, a simple machine to move sheet metal past an array of noncontact sensors was designed and assembled. This machine consists of two 25 cm diameter drums connected together by a continuous belt of the sheet metal to be investigated. A variable speed motor drives one drum to determine the speed of the metal belt while position-adjusting screws on the other drum establish the tension in the belt. An open area of about a square meter above and below the belt is available for mounting sensors that can interact with the sheet across air gaps of appropriate size. Initial tests of the machine will be conducted in early FY96.

Task II. Precision Ultrasonic Wave Velocity Measurements. The techniques available for making precise measurements of ultrasonic wave velocity usually require bulk samples that have flat, parallel surfaces machined onto them for determining transit time or have regular polyhedron shapes that can vibrate at well defined resonant frequencies. For thin sheet metal, through-thickness-resonance methods can be used to accurately determine the ultrasonic wave velocities in the thickness dimension. Unfortunately, the desired commercial properties of sheet metal are usually attained by developing a preferred orientation in the grains that make up the microstructure. This texture makes it necessary to measure the ultrasonic wave velocities in directions other than the thickness dimension. Sound waves that propagate in the plane of the sheet are Lamb waves, and their velocities not only depend on the frequency being used but they have complicated, nonlinear relationships with the elastic constants of the constituent grains. Fortunately, the electromagnetic acoustic transducer (EMAT), whose noncontact characteristic makes it the transducer of choice in a manufacturing facility, can easily be designed to excite and detect only one Lamb wave mode at a well defined frequency. This simplifies the experimental procedures and the mathematical analysis so that phase velocities can be determined very accurately and the mathematical relationships with the elastic properties of the sheet can be written down in closed form and accurately solved by a computer.

During FY95, several types of EMATs were designed, built and tested in fixtures that allowed the phase velocities of various Lamb-wave modes to be measured to precisions approaching ± 0.1%. Both shear and extensional types of waves were measured as a function of angle relative to the rolling direction in the plane of the sheet. This, in principle, provides many independent measurements that will be used to unravel the complicated expressions relating sound velocities and elastic constants of crystallite.

Task III. Correction for Texture Effects. The symmetry of the rolling process forces an orthorhombic symmetry onto the physical properties of the sheet. Thus, a complete description of the elastic or ultrasonic properties of the sheet requires specification of nine elastic modulus tensor elements. Fortunately, theoretical analysis of the properties of polycrystalline aggregates has been developed to a high level during the past several years. Now it is possible to find mathematical expressions in the literature that relate the nine elastic moduli of a textured sheet to three orientation distribution coefficients (ODCs) that describe the texture and the three single-crystal elastic constants. These describe the elastic response of an individual grain if that grain is a crystal with cubic symmetry. The more recent literature also contains mathematical expressions for the relationships between certain Lamb-wave velocities and the nine tensor elements of the elastic modulus. Some of these are the result of series expansions in small quantities, so care must be exercised in applying them to specific cases. During FY95, the expressions given in the literature were used to develop a mathematical inversion procedure for deducing the three ODCs and the three effective single-crystal elastic constants for a grain from measurements of Lamb-wave phase velocity in the plane of the sheet and of bulk wave velocities in the thickness dimension. All of the velocities can be measured by using EMATs that operate across an air gap over moving sheet metal.

Task IV. Applications. The most simple application of the experimental and theoretical techniques developed to date is to rolled brass and copper, where the texture effects are large but the grains are single-phase cubic crystals. As part of a CRADA with Olin Corporation, EMAT techniques are being developed to monitor recrystallization and grain growth during the rolling and annealing processes in an operating mill. Since Olin has supplied sheet samples of copper, bronze, and brass with various thicknesses, degrees of rolling reduction and annealing histories, the procedures to be used for following the development of texture during recovery, recrystallization, and grain growth in the rolling mill are becoming established.

Another simple case for on-line monitoring of sheet-metal properties is being investigated as part of an ATP project with Allied Signal. Here, the analysis is simplified because the material is METGLAS, which is an elastically isotropic metallic glass. Thus, there is no texture, and only two elastic constants are sufficient to describe the elastic response of the material. Measurements of the phase velocities of the So and SH Lamb wave modes were sufficient to define these two elastic constants to an accuracy of ± 0.1%. Since Allied Signal supplied samples with different processing histories and different ductility properties, it was possible to seek correlations between the measured wave velocities and the mechanical strength property of ductility. A linear correlation between the ductile-to-brittle transition temperature (DBTT) and the product of the SO and SH wave velocities was observed; a model to explain this is being developed.

A third study addresses the application of Lamb-wave velocity measurements to steels. Here, we are taking advantage of an AISI/DOE program at NIST to develop constitutive relations that describe the process of hot rolling of steel in the austenitic phase. This program also seeks to develop quantitative relationships between the mechanical strength properties of the final product and the microstructures developed after the cooling of the steel through the gamma-to-alpha phase transition and the formation of martensite. EMATs that use magnetostriction as the coupling mechanism are being developed to measure both the Lamb-wave velocities and the magnetic properties of these steels at the same time. Ultimately, this will provide both magnetic and acoustic measurements of physical properties to compare with the mechanical properties and microstructure information being developed for the steels of interest to the AISI and DOE.

FY95 Outputs

1. Proprietary letter reports to Olin Corp. and Allied Signal.