Effects of spot welding on the functionality of thin-wall elements of car's deformation zone/Suvirinimo kokybes itakos automobilio deformaciniu zonu elementu funkcionalumui tyrimas.

Juodvalkis, D. ; Sapragonas, J. ; Griskevicius, P. 等

1. Introduction

Our roads are filled with second-hand vehicles. Needless to say, that those cars frequently happen to be used after some accident. Almost every accident of a car leads to irreversible damage in the passive safety systems of the vehicle. Fortunately some of those systems can be replaced or repaired, for example, safety airbags with their control modules and seat belts with tension sensor assemblies can be replaced after the car crash. However the major problem is with the car body and its deformational zones.

Car crash under higher speeds usually leads to substantial changes in the body geometry and damages of deformational zones [1-3]. As it often happens, geometry of the car body and deformational zones are repaired after the accident making it possible the further use of the car. Repair of deformational zones includes straightening, sometimes welding, and some parts get simply replaced [4, 5]. In any case such restructuring changes mechanical properties of materials as they are exposed to plastic deformation and are thermally processed. This changes characteristics of some junctions too, as instead of factory spot-welding parts are inter-welded manually.

The aim of this research is to estimate the influence of the spot-welded connections quality on the main crash characteristics as absorbed energy, deformation form or axial shortening. The experimental quasi-static and impact tests of the axial crushing of columns were performed to validate finite element (FE) models which were used for the simulations to evaluate the influence of spot-welding.

2. Experimental research and results

The U-shape specimen was selected for the purpose of our research. Thin-walled columns with the length of 20 cm had the similar cross-section as automotive longerons (Fig. 1).

The experimental quasi-static and impact tests were performed with the thin-walled columns (Fig. 2). The layout of spot-welding was symmetrical.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

The quasi-static axial compression tests were run on the universal hydraulic 5 t tension-compression testing machine, which applied the axial load through flat end platens without any additional fixing. Crosshead speed was approximately 0.3 mm/s.

The mean crush load can be calculated by Wierzbicki and Abramowicz [6] analytical equation

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (1)

where [[sigma].sub.0] is the mean plastic stress, t is the thickness, b is the width of the longerone.

In the test the specimen is buckled at crushing distance of 100 mm. The load-compression curves were obtained.

The dynamic testing equipment (Fig. 3) involved impact by the solid of 60 kg in weight with the initial speed at the impact of 6 m/s. The required speed was obtained by selecting a height from which the mass was dropped down. To prevent the impacting mass from deviations and to ensure the symmetric impact, a specific guiderod 4 with the diameter of 10 mm was installed. With the purpose of a very precise placement of a specimen, the specific centring components 2 were used. Acceleration sensors 6 were installed on the impacting mass. Accelerometers of Type "Wilcoxon 784 A" were used for testing. Characteristics of accelerometer: Sensitivity, 25[degrees]C 100 mV/g Acceleration Range 50 g Frequency Response 4-7000 Hz Power required voltage 18-30 V Sensing Element Design PZT ceramic Weight 45 grams

The analogue signal of sensors readings was converted by Device "Picoscop - 3424" to the digital signal that was recorded on a computer using the program "Picoscope 6" designed by the manufacturer of this device. Several initial experiments lead to the observation that in the initial stage of the impact, high-frequency vibrations occurred that were undesirable in our case as they significantly distorted findings of our testing. To reduce the vibrations a mechanical damper was used between the specimen and the impacting mass.

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

To compare the simulation results with the experiment of axially crashed thin-walled elements, it is important that the same material properties as used in the experiment are also used in the material model of the simulation [7, 8]. 6 plates (20x300 mm) are selected as samples for the test. To approximate the tension diagram the material model *MAT_PLASTIC_KINEMATIC in LS-Dyna [9, 10] with linear isotropic strain hardening approximation was used:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (2)

where E is Young's modulus; [[sigma].sub.Y] is yield strength; [[epsilon].sub.Y] is yield strain; [E.sub.t] is tangential modulus

Tangential modulus was calculated

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (3)

where [[sigma].sub.v] is true ultimate tensile stress; [[epsilon].sub.v] is true uniform strain

The main mechanical properties of the material used for FE simulations are presented in the Table 1.

The quasi-static experiment, dynamic crushing experiment and FEA simulating results of the longerone deformation histories are shown in Fig. 5 and Fig. 6.

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

3. The influence of spot-welded connections to the parameters of axially crushed columns

In automobiles, deformation zones are usually made of several tinplate sheets that initially undergo pressing to obtain their specific form. These thin-plate sheets are interconnected by spot-welding. When repairing sometimes it is necessary to separate individual panels of the deformation zone in order to recover their primary form (Fig. 7, a, b and c). The most common way for separating the spot-welded sheet metal parts is drilling out using the special drilling device. The diameter of the drilling device is selected according to the diameter of the spot-weld. Predominant diameter of the drill is 8 mm. After recovering the primary form of the panels they are interwelded again. The panels are welded manually at the former spot-welded areas.

[FIGURE 7 OMITTED]

In case of manual welding the strength of spot-welds is dependent on the following factors: preparation of the connective components (elimination of paints, primers and corrosion), intensity of the welding current and of course the level of qualification of the welder [11]. Selection of inadequate manual welding mode can result in a welding of a very low quality. At least one missed spot-welding doubles the spacing between adjacent connections in our research profile.

Critical axial and shear forces are the main mechanical characteristics of the spot-welded connections. The experimental testing of manufactory and manually spot-welded connections (Fig. 7, d) were performed to determine the axial and shear forces that destroy the connection. Table 2 presents the results of the test.

Experimental testing of the strength of spot-welded connections showed that the critical axial force at the manually welded spot-welds is reduced to 4.2 kN. Critical shear force remains almost constant.

The spacing of mechanical connections between components, to develop a composite section, is typically based on several strength considerations. The spacing between the spot welded joints according to AISI recommendations should be chosen to prevent Euler buckling of the more slender component at the desired stress level.

Slenderness of the full section is calculated considering that the effective length factor of axially compressed column is equal to 0.5 [12].

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (4)

where r is radius of gyration; [L.sub.ef] is effective length of the column, I is principal moments of inertia, A is area of cross section.

Slenderness characteristics of separate elements are shown in Table 3.

AISI requires a connection spacing that limits the slenderness ratio of the channel between connectors to the critical slenderness ratio of the composite section. For connected component the effective length factor is considered at 0.6 [12].

The spacing between spot-welded connections is calculated:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII.] (5)

where [[lambda].sub.sum] is slenderness of full column, [r.sub.min] = 5.21mm is minimum radius of gyration of separate component.

Dynamic experiments lead to the observation that axial shortening was significantly larger in those specimens that had some ruptured spot-welds although load (mass and initial speed) was the same in all the cases. The specimen with two ruptures of spot-welding was deformed by 11 cm, whereas the specimen with no spot-welding ruptures was deformed only by 9.5 cm (Fig. 8).

[FIGURE 8 OMITTED]

[FIGURE 9 OMITTED]

Experimental results of axially crushed thin-walled elements with factory and manually spot-welded connections are shown in Fig. 9 and 10.

[FIGURE 10 OMITTED]

Based on the findings of above-described experimental tests and results of calculations, we have examined, using FE model, the influence of partial or total failure of a spot-welding on the deformation of the specimen and amount of the energy absorbed. Fig. 11 presents the layout and numbering of spot-welding.

[FIGURE 11 OMITTED]

Now, let's investigate the influence of impaired spot-welds on the deformation of the specimen. Variations of manual and original spot-welding by FEA simulation are shown in Table 4.

Fig. 12 shows the influence of spot welded connections on the deformed shapes, their strength characteristics and distance between the spot welding.

It is obvious that greater distance between spot-welded connections may have significant influence on the axially crushing behavior and therefore can significantly decrease the amount of energy absorbed. Using characteristics of axial strength of the manually spot welded connections the deformed form looks similarly except for the fact that the absorbed energy decreased slightly.

The typical force deflections curves of axially crushed thin-walled columns obtained from FE simulations are presented in Fig. 13.

[FIGURE 12 OMITTED]

Connections welded by manufactory or manually in the FE model have been evaluated changing the data of normal and shear failure forces. Taking in to consideration that in some cases of repairing process we could obtain faulty welded connections in FE model it is simulated by removing spot welds. Comparing the curves of axially crushed columns connected by factory welding, manual welding at 5, 6 positions and nonwelded 5, 6 positions we can see, that at the beginning the force-deflections curves are similar in all the cases. Later, at the moment of second wave formation the differences are visible. The columns with nonwelded 5, 6 positions do not have second wave therefore the energy absorption is significantly smaller than in other cases. The columns with characteristics of manually welded connections initiate the formation of second wave but need smaller axial force for it therefore the energy absorption is slightly smaller.

[FIGURE 13 OMITTED]

4. Conclusions

1. Comparing the experimental results of statically and dynamically axial crushed thin-walled columns we see that the dynamically deformed thin-walled columns absorb from 10 to 50% more energy than those statically deformed.

2. The results of FE simulations show good match with the results of dynamical (drop weight) experiments, therefore the influence of spot-welded connections characteristics on the behavior of axially crushed thin-walled columns were analyzed using FE simulations.

3. By FE simulations it was determined that:

* depending on the specific position, characteristics of spot-welding can have more or less influence on the amount of energy absorbed at one unit of length;

* in case of the axial shortening of 0.092 m, failure of the spot-welds in those positions where a wave forms at the initial moment causes the energy absorbed by the specimen to reduce by 15%;

* comparing the behavior of axially crushed columns connected by factory welding between the columns with nonwelded connections in the middle of them we can see, that at the beginning the force-deflections curves are similar but at the moment of second wave formation the differences are visible. The columns with nonwelded middle do not have second wave therefore the energy absorption is significantly smaller than in other cases.

Received April 01, 2009

Accepted May 27, 2009

References

[1.] European Vehicle Passive Safety Network 2, Task 4.6. -Vehicle Structural Crashworthiness, TNO Automotive Crash Safety Centre, 2004.-15p.

[2.] Juodvalkis, D., Sapragonas, J., Griskevicius, P., Ostrowski, M. Investigation of cars' damaged deformational zones characteristics.-Transport Means-2008, proceedings of the International Conference.-Kaunas: Technologija, 2008, p.245-249.

[3.] Griskevicius, P., Ostrowski, M., Holnicki-Szulc, J., Zeleniakiene, D. Feasibility study of vehicle adaptive front structure.-Proceedings of the International Conference "Transport Means", 2006. ISSN 1822-296 X. -Kaunas, 2006, p.289-293.

[4.] Advanced High Strength Steel Repairability Phase 1. American Iron and Steel Institute. Final report. 2000. -22p.

[5.] Robinson A., Livesey W.A. Repair of Vehicle Bodies. -Oxford, 2000.-626p.

[6.] Abramowicz, W., Jones, N. Dynamic progressive buckling of circular and square tubes. -International Journal of Impact Engineering, 1986, v.4, p.243-270.

[7.] Bagmutov, V., Babichev, S. Features of stress strain state in specimen neck at (when) computationally modeling a tension process. -Mechanika. -Kaunas: Tech nologija, 2005, Nr.5(55), p.5-10.

[8.] Januteniene, J., Didziokas, R., Gintalas, M. Analysis of the variation of metals mechanical properties depending on operation time. -Mechanika. -Kaunas: Technologija, 2009, Nr.1(75), p.26-30.

[9.] Leisis, V., Sudintas, A., Ziliukas, A. Prediction of the strength and fracture of the fuel storage tank. -Mechanika. -Kaunas: Technologija, 2008, Nr.4(72), p.14-20.

[10.] Hallquist, J.O., "LS-DYNA Theory Manual", Livermore Software Technology Corp., Livermore, 1998. -498p.

[11.] Griskevicius, P., Leisis, V., Zeleniakiene, D. Impact Energy Absorption of Structural Elements.- Mechanical Technology XXXIII. Scientific works ISSN 16484711. -Kaunas: Technologija, 2005, p.30-37.

[12.] Automotive Steel Design Manual. American Iron and Steel Institute & Auto/Steel Partnership, Revision, 2000, v.6.-747p.

D. Juodvalkis *, J. Sapragonas **, P. Griskevicius ***

* Kaunas University of Technology, Kqstucio 27, 44312 Kaunas, Lithuania, E-mail: djuodvalkis@yahoo.com

** Kaunas University of Technology, Kqstucio 27, 44312 Kaunas, Lithuania, E-mail: jonas.sapragonas@ktu.lt

*** Kaunas University of Technology, Kqstucio 27, 44312 Kaunas, Lithuania, E-mail: paulius.griskevicius@ktu.lt Table 1 Mechanical properties of the material Nr. [[Sigma].sub.y], MP [[sigma].sub.Q], MPa [[epsilon].sub.u] 1 287.2 430.1 0.262 2 286.4 424.7 0.270 3 306.6 430.1 0.255 4 306.9 424.7 0.266 5 296.4 421.9 0.259 6 296.6 416.7 0.261 av. 296.7 424.7 0.262 CV 3.01 1.20 2.01 Nr. [E.sub.t], MPa 1 547.4 2 514.6 3 487.9 4 445.0 5 488.3 6 462.7 av. 491.0 CV 7.44 CV is coefficient of variation Table 2 Mechanical properties of spot-welding connections Nr. Original spot-welding Shear Axial [F.sub.max] * kN [F.sub.max] * kN 1 9.8 7.6 2 10.2 9.2 3 9.6 8.0 4 9.8 7.8 5 10 8.4 6 9.6 8.2 av 9.8 8.2 CV 2.41 6.89 Nr. Manual spot-welding Shear Axial [F.sub.max] * kN [F.sub.max] * kN 1 9.6 3.8 2 9.4 4.2 3 9.5 4.3 4 9.6 4.6 5 8.6 4.0 6 9.1 4.4 av 9.3 4.2 CV 4.19 6.82 CV is coefficient of variation Table 3 Slenderness characteristics of separate elements Principal moments Nr Area A, of inertia Radius of gyration [mm.sup.2] [I.sub.x] [I.sub.y], [r.sub.x], [r.sub.x], [mm.sup.4] [mm.sup.4] mm mm 1 144 62382 114209 20.81 28.16 2 74 2011 41093 5.21 23.56 Sum 218 100403 155302 21.46 26.69 Nr Slenderness of elements [[lambda].sub.x] [[lambda].sub.y] 1 4.80 3.55 2 19.18 4.24 Sum 4.66 3.75 Table 4 Variations of manual and original spot-welding by FEA simulation Axial Absorbed Nr. Spot-welding shortening energy * F * Original Manual m % kJ % kN 1 1-10 -- 0.092 100 1.16 100 12.6 2 1,2,5-10 3,4 0.116 126 0.99 85 10.8 3 1,3,5-10 2,4 0.105 114 1.02 88 11.1 4 1-4,7-10 5,6 0.103 112 1.07 92 11.6 5 1-4, 9, 10 5-8 0.103 112 1.07 92 11.6 6 1,2,7-10 3-6 0.118 128 0.99 85 10.8 7 1-6 7-10 0.119 129 0.99 85 10.8 8 1,2 3-10 0.119 129 0.99 85 10.8 9 -- 1-10 0.122 133 0.97 84 10.5 10 1,2,4-10 3 ** 0.106 115 0.97 84 10.5 11 1,2,5-10 3,4 ** 0.129 140 0.87 75 9.5 12 1-4,7-10 5,6 ** 0.147 160 0.83 72 9.0 13 1,2,4,6-10 3,5 ** 0.147 160 0.83 72 9.0 14 1,2,7-10 3-6 ** 0.158 172 0.76 66 8.3 15 1-4,9-10 5-8 ** 0.149 162 0.82 71 8.9 16 1-6,9,10 7-8 ** 0.124 135 1.00 86 10.9 *--with 0.092 m axial shortening. **--not connected spot-weld locations (Faxcjal = 0N).