Competing or complementary joining methods?
February 12, 2008
Projection welding, also known as resistance fastener welding, is the most common method for nut and M- or T-thread bolt welding. This article examines recent experiments with projection welding of those components to various high-strength steel components.
Weld Nuts Studied
A modern car body contains some 300 welded and punched fasteners, such as bolts, nuts, and studs. The quality of the attachment of these fasteners to the stamped body components is critical for the final product's safety and reliability. Crucial components such as the front and rear axles are mounted to such fasteners, the seat belts and steering column are anchored to them, and they provide grounding for electrical wires.1
Projection welding is the most common method for nut and M- or T-thread bolt welding. These nuts and bolts are provided with annular projections or three to four separate projections, depending on the application. With an annular projection and a plastic insert, some nut types can achieve a watertight joint. The annular projection also is preferable for attachments subjected to fatigue stress and high loads. Compared to fasteners with separate projections, the annular projection offers fewer crack initiation points.
Steel Sheet Materials Studied
Resistance fastener welding, generally called projection welding, is similar to resistance spot welding. However, in the spot welding process, the size of the contact surface of the electrode cap tip determines the current flow, whereas in projection welding, the current flow is constricted to the embossed or machined projection. Both AC and DC power sources are suitable for fastener welding. The heat balance for projection welding is affected by the following factors:2
Projection welding of a weld nut or weld bolt involves three phases. In the first phase, the projection is in contact with the mating sheet. Then the current starts to heat the projection to welding temperature. The electrode force then causes the heated projection to collapse rapidly, and fusion takes place.
The current for projection welding is generally less than that required to produce correspondent spot welds. The projection will heat rapidly, and excessive current will melt it and result in expulsion. However, the current must be at least high enough to create fusion before the projection has completely collapsed.
A short welding time might be desirable from a production standpoint, but it will require correspondingly higher amperage. Therefore, it is important to optimize welding parameters to prevent overheating and metal expulsion. In some cases, such as when welding fasteners to high-strength steels (HSS), impulse welding may be advantageous to control the heating rate. This also is helpful for thick-sheet projection welding and when welding metals with low thermal conductivity.
The electrode force should be adequate to flatten the projections completely when they reach welding temperature and to bring the fastener in contact with the sheet metal part. Excessive force will prematurely collapse the projections, and the weld will have an incomplete fusion in the center. The welding machine must be able to follow the movement of the electrodes as the projections collapse. Slow follow-up will result in metal expulsion before the parts have been brought together.
Three weld nut types were selected for the trials (seeFigure 1). Two thread sizes (M6 and M8), commonly used in body-in-white (BIW) manufacturing, were investigated. The sheet materials used in this study (see Figure 2) were from 460-megapascal (MPa) to 1,550- MPa ultimate strength.3
The welding trials were aimed at establishing the largest possible current range, from minimum approved torque value—32 newton meters (Nm) for M6 and 63 Nm for M8—of the weld to the spatter limit. Only single-pulse welding was investigated with fixed parameter settings (see Figure 3). Stationary resistance welding machines with both MFDC (middle frequency direct current ~1,000 Hz) and AC (alternating current = 50 Hz) were used for the welding trials, and external measurement devices were used to monitor the welding process.
General Welding Parameters for the Study
Figure 4shows the pull-out load as a function of the welding current for the different weld nuts welded to Dogal 600DP (1.5 mm thick). The colored lines represent the approved current range for each nut type.
Pull-out loads as a function of welding current for different nut types welded to 1.5-mm Dogal 600DP utilizing both MFDC and AC are shown. The highest and lowest welding current values represent the boundaries in the approved current range.
Because of the larger projections of the M8 nuts, higher welding current was needed to create a weld, compared to the M6 weld nuts. Moreover, the welding current was dependent on the geometric design of the projections. The annular projection of the high-performance (HP) weld nut required much higher welding current compared to the three and four projections of the hexagonal (Hex) and square (SQ) nuts, respectively.
As expected, the highest pull-out force was recorded for the HP M8 weld nut for all sheet materials in the study. It is, however, interesting to note the surprisingly large difference in strength between nuts welded with AC and those welded with MFDC. The AC-welded nuts showed a superior pull-out force. This is explained by the much more rapid current built up with AC welding, which improves the formation of the weld. AC also increased the current ranges for SQ and Hex nuts.
To improve the nut performance and joint strength with MFDC welding, trials were performed with different prepulses and current ramp settings. Applying a very high initial current pulse followed by a lower one improved the pull-out strength of a Hex M6 nut by more than 100 percent. Despite the current shape modification, however, the pull-out forces were always lower compared to nuts welded with AC.
The loading of the weld nuts in the tests does not represent how the loads are applied in real body applications. Here the nut is mounted on the back side of the sheet (see Figure 5), which generally is called a nut plate and normally is of thicker material than the surrounding sheet components. In this manner, the nut has to be pulled through the sheet at extreme loading conditions, something that, of course, increases the strength of the overall screw joint significantly.
The different loading conditions for welded nuts are shown. Pull-through, on the left, was used in the strength validation tests, and pull-out, on the right , was used during process optimization trials.
Three types of M8 threaded weld nuts—SQ with four separate projections, Hex with three separate projections, and round HP with an annular projection—were included in the pull-through tests.4The sheet material was zinc-coated DP600 in three thicknesses: 1.2, 1.65, and 2.0 mm.
The sheet metal was quadratic, measuring 125 by 125 mm. A hole with a diameter of 10 mm, relative to which the nut was centered before welding, was punched in the middle of the sheet. Five repetitive tests were done for each combination.
In each test a "fresh" bolt, specified as M845 12.9 DIN 912 with socket head cap type, was threaded into the nut. The load then was applied to the socket head by pull-through loading using a servo-hydraulic tensile test machine and a testing speed of 10 mm per minute.
The number of different failure modes from pull-through testing in D600 are listed.
In the pull-through tests, sheet failure occurred in almost all the trials for the 1.2-mm- and 1.65-mm-thick sheet, whereas thread failure was the most common failure mode for nuts attached to the 2.0-mm-thick material (seeFigure 6). The influence of sheet thickness on the different nut types can be seen in Figure 7.
Click to view image larger This chart compares the average maximum stresses from pull-through testing of square, hexagonal, and high-performance weld nuts attached to DP600 materials in 1.2-, 1.65-, and 2.0-mm thicknesses, respectively.
The maximum stress calculated on the Y axis was based on the total contacting flange surface of the weld nut, not only the welded area.
The influence of sheet thickness on the maximum load was largest for the Hex nuts. Increasing the sheet thickness from 1.2 to 1.65 mm (37 percent) increased the maximum load by 21 percent, and increasing sheet thickness from 1.65 to 2.0 mm (66 percent) increased the load by 43 percent.
The tendency for thread failure in 1.65-mm-thick material was larger for the HP nuts because of a larger contact flange area—7 percent larger than Hex and 26 percent larger than SQ.
The strength and thickness of the sheet material and the contact area between sheet and weld nut flange affect the failure mode and the maximum load. With thicker sheet, the screw joint was more prone to thread failure.
The maximum displacement, measured at the point where the load starts decreasing, was mainly influenced by the failure mode. The displacement always was shorter when thread failure occurred.
Weld Nut Specimen Configurations for Fatigue Testing
The sheet metal fatigue life close to a weld nut is affected by the type of nut and its dimension, the sheet material thickness and strength, the type of loading, and the load direction.
Critical weld nut or bolt joints in the car body structure are subjected mainly to bending fatigue loads, which initiate fatigue cracks in the sheet metal. Therefore, a fatigue experiment was conducted to evaluate sheet metal fatigue characteristics during bending fatigue loading conditions.5
The experiment included variation in nut geometry and dimension, as well as sheet metal grade and thickness. Both single- and double-sheet testing was performed (see Figure 8). In the latter configuration, the nut was welded to the thicker sheet, which presented the highest yield strength.
Two of the three weld nut types used in the static testing also were used for fatigue evaluation: the Hex with three separate projections and the round HP nut with an annular projection. The study also included a comparison between M8 and M10 threaded nuts.
Summary of Results From Fatigue Experiments
The sheet specimen was manufactured as a circular plate with an outer diameter of 120 mm. A hole was punched and a nut was projection-welded to the center of the plate. The test object was put in a fixture, where the circular clamping guaranteed a free sheet diameter of 100 mm.
A coupling was fastened to a socket where the screw joint was preloaded according to standards. The fatigue load (force) was applied to the test object through this coupling, where the length of the lever arm (thick pipe) was 70 mm, to generate bending of the plate around the X axis.
The results from the fatigue experiments are summarized in Figure 9. The scatter in the results was considered low since each data set was more or less on the curve fitted to it.
The HP nuts proved less damaging compared to the Hex nuts for the same magnitude of reversed loading, and the HP nut configuration survived more load cycles.
An increase in sheet metal thickness seemed to be the best way to improve the fatigue performance of a welded nut, whereas an increased thread size or the introduction of a higher-strength sheet material resulted in only minor improvements. Introducing a high-strength steel for weld nut attachments subjected to small-amplitude fatigue loads is not recommended.
In the double-sheet specimen tests, the fatigue crack always started off and grew in the thicker sheet to which the nut was welded. This test resembled more the real loading conditions that appear in a BIW attachment and clearly demonstrates the load distribution effects in a double-sheet configuration, which helps improve the fatigue performance of the welded nut.
Many factors influence the choice of fastener type, such as loading conditions, sheet thickness and grade, manufacturing preferences, quality assurance, and function. It also is important to establish which part in the screw joint system is the limiting factor (thread strength or the fastener's attachment to the sheet part) to avoid expensive overdimensioning.
For lower loads and small thread sizes (M6 and M8), punch and weld fasteners could be considered equal in their ability to carry the load. In such cases, characteristics of the specific manufacturing process might determine which fastener is best.
For higher loads and larger thread sizes, a welded fastener is recommended. For these loading conditions, a weld nut also should be localized on the sheet side opposite from the loading direction. This design minimizes the influence of weld defects because the nut has to be pulled through the sheet or nut plate. The thickness of such a nut plate can be maximized to 2.0 mm. Higher gauges make no sense, because the thread strength then seems to be the limiting factor.
The introduction of more high-strength steels in car bodies also has affected fastener technology. Because of the alloying concept of these materials, welded fasteners have problems with reduced weldability. AC seems to be superior to DC welding in terms of presenting a broad current range or welding lobe. However, with an appropriate weld parameter setting, such as pulse shaping, even highly alloyed grades, such as TRIP (transformation-induced plasticity) and boron-alloyed steels, can be welded with satisfactory quality.
For crash loads and fatigue loads in the short life regime, fasteners with annular projections are a good choice, although they also are the most difficult to weld. A flat sheet surface is necessary to meet the continuous projection. Also, no reliable nondestructive test method is available for this type of weld nut. Therefore, quality assurance involves destructive checking on a subassembly level to prevent high scrap rates and expense.