Abstract:

In the plasma arc welding process, the current used and the speed at which the torch moves are two important factors in determining the properties of a weld. We make a number of plasma welds using different currents and speeds. Real-time temperature data is taken and analyzed using the current mathematical model. Cross sections of the welds are inspected with a microscope. We attempt to use a mathematical temperature model to explain the different structural regions of the weld seen under the microscope.

Introduction:

The plasma arc welding of stainless steel is a vital part of many industrial processes. Two of the most important factors in determining the properties of a plasma arc weld are the speed with which the weld is made and the amount of current used to make the weld. A better understanding of how these two conditions affect plasma arc welds could lead to more efficient welding processes. In this project we will create a database of welds which were made at various speeds and currents, keeping all other factors constant. We will then be able to analyze the physical properties of the welds to determine the affects of both speed and current on the final weld properties. This database of welds can be used to verify the accuracy of a MathCad program that will numerically model the welding process and predict weld parameters based on a known weld speed and current.

In previous work an incomplete database of welds was compiled and organized. Our intention is to fill in all gaps that exist in this file as well as extend it, and perhaps make repeats of certain welds to make comparisons. The table below shows images of the cross-sections of the previously completed welds.

Table 1: Previous year’s collected database of plasma welds

The temperature of the samples during the welding process is also of interest, as temperature affects the micro structure of the metal and therefore its properties.

Temperature data can be taken directly, using thermocouples, or indirectly with an infrared camera. In our experiments we will use both methods and compare the results to determine which is the most accurate and efficient method of recording temperature data.

Experimental Setup and Design:

Using a Staubli plasma arc welder we welded 1/8" x 1" x 3" samples of stainless steel at various speeds and currents. The length of the plasma arc was kept constant at 2mm. We recorded the welding process using an Inframetrics 600 infrared camera, creating images similar to the following one.

Figure 1: IR Camera Image Figure 2: Thermocouples ultra-sonically

welded to stainless steel sample

For several of the samples we also welded four thermocouples leads around the welding path using an ultrasonic welder. These thermocouples were fed into a Hydra Fluke Data Acquisition Unit and temperature vs. time data was gathered during the welding process using a LabView program. These thermocouple readings were then compared to the temperature readings taken from the infrared recording at the same point.

Figure 3: Entire laboratory set-up including: robot, infrared camera, fluke hydra, television and computer.

After all samples were welded, they were cut at a point where the weld was consistant for cross-sectional inspection. The samples were placed in self-hardening clear plastic molds for easier viewing under the microscope. The molds and cross-sections were polished in a five step process, down to 0.05 micron tolerance, to remove all scratches. Finally the cross sections were chemically etched to make the metal structure visible.

visible weld

Figure 4: Cross section of welded stainless steel sample, mounted in plastic mold

Samples were inspected under a microscope. The width, depth and varying structural regions of each individual weld were recorded.

Results:

Discussion:

Conclusion: