Page images
PDF
EPUB

AN INTEGRATED FEA BASED PROCEDURE FOR WELD FIXTURE DESIGN

Z. Yang, X. Chen, Y. Dong

E. Martin*, D. Michael*

ABSTRACT

As mechanical forces and welding induced thermal stress are applied during welding, the strength of the weld fixture is one of the crucial factors determining a successful or unsatisfactory welding practice. In the present study, an integrated FEA based procedure for welding fixture design was established. In the integrated FEA procedure, the welding induced stress and distortion were simulated by sequentially coupling thermal and mechanical modeling using a comprehensive welding simulation package. The welding simulation results showed that thermal induced reaction forces were significant at the constraint locations, which must be considered for weld fixture design. The magnitudes of reaction forces varied depending on constraint locations during welding process. The obtained maximum reaction forces were then input into weld fixture structural analysis. The predicted stress and distortion from the fixture structure analysis provided a guideline for the weld fixture design.

KEYWORDS

Welding Simulation, FEA, Stress and Distortion, Fixture Design

INTRODUCTION

Residual stresses and distortions are two of the major concerns in welded structures. The welding induced residual stress can exceed the yield strength of the material and is detrimental to the integrity and fatigue life of the welded parts. Welding can also cause the welded products to distort significantly from their original shape, which may require costly post-weld treatment such as machining or straightening. In recent years, tremendous efforts have been made to reduce weld distortion. Various methods have been proposed to accomplish this, including precambering, pre-bending, preheating, and thermal tension.

Thermal-mechanical FEA of the welding process is an emerging and rapidly maturing technique. Computer aided design of the welding process is becoming an efficient and effective approach to achieve high quality weld products in industry (Ref. 1-5). Welding simulation helps to optimize welding procedure (welding parameters, sequence, and weld joint geometry) and apply appropriate mechanical or thermal methods to reduce welding induced residual stress and distortion. In recent years, welding computer models have demonstrated the capability to reduce fabrication costs, improve weld quality and increase service durability by optimizing the weld process.

Technical/Service Division, Caterpillar Inc., 1311 East Cedar Hills Dr., Mossville, IL 61656-1875
Mining and Construction Equipment Division, Caterpillar Inc., 27 Pershing Rd., Decatur, IL 62525-1817

The weld fixture is normally used to facilitate assembly of parts and to hold the parts in a fixed relationship for welding fabrication. In welding industry, it is common to employ techniques prior to the welding processes that induce distortions in the material that are essentially opposite to the distortions induced by the welding process. Pre-cambering is such a method that places the material into a weld fixture, which holds the material in the desired distorted shape until welding is completed. The intent of the induced distortions is to mitigate the distortion induced by welding to maintain the desired shape after welding.

Weld fixtures have typically been designed based on empirical data and past experience. Many trial-and-error tests are needed for designing a new weld fixture, which leads to not only costly design but also delays in production. Furthermore, a weld fixture for pre-cambering must be designed to withstand reaction forces from the mechanical distortions induced on the material, and from thermal distortions produced by the weld process. However, the thermal induced reaction forces were usually ignored or not fully understood in previous fixture design due to the lack of knowledge in this area. As a result, the designed weld fixture might deform or even fracture during welding, which significantly affected the quality and reliability of welded product.

In this paper, an integrated FEA based procedure for weld fixture design is proposed. In this procedure, the thermal and mechanical reaction forces were considered in combination to design weld fixtures for welding processes, especially for pre-cambering technique. Welding induced distortion and thermal reaction forces were calculated by using a comprehensive welding simulation tool (Ref. 1-5), which has been proven to be very efficient and effective in prediction of weld distortion and residual stress in large fabricated structures in industry.

1. Modeling Algorithm

OUTLINE OF STUDY AND METHODS

The algorithm of the integrated FEA based fixture design procedure is shown in Fig. 1. The procedure can be categorized into three steps: (1) model preparation, (2) welding simulation, and (3) fixture strength analysis. Steps (1) and (3) are common in FEA for structure analysis. Step 2, welding simulation, is critical in the present study. It calculates the thermal distortion in the

[blocks in formation]

weld products and reaction forces to the fixture during welding process. The reaction forces were then input into weld fixture model for stress analysis and distortion prediction in order to examine the stiffness of the designed weld fixture.

1.1 Welding Thermal Analysis

The welding simulation was done by sequentially coupling thermal and stress analysis. Two approaches, numerical or analytical solution, can be taken for thermal analysis. Numerical solution is the commonly used approach in FEA thermal-mechanical analysis. However, it requires extensive manpower and computer resources to analyze complex weld structures. In the present study, an analytical based Comprehensive Thermal Solution Procedure (CTSP) was used, which is able to simulate temperature-time history for various types of weld joints (V-groove, Tfillet and lap joint) in complex welded structures (Ref. 1, 2). The details of CTSP are given in literature and only its essential features are given here (Ref. 1):

[merged small][merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small]

where T(,y,z,t) is the transient temperature field, T,(,y,z,t) is the quasi-static state temperature field, F(5,y,z,t) is the transient transform function, Q is welding heat input, & is conductivity, R is the radius from point heat source, v is welding speed, a is thermal diffusivity, erf is error function, and = x - vt, with the heat source traveling in the x direction. It should be noted that equations (1), (2), and (3) are valid for an infinitely thick plate. For plates with finite thickness, it is assumed that the heat loss from the surfaces is negligible in comparison with that due to conduction in an infinite body and a number of reflected heat sources are used to balance heat loss at surfaces.

1.2 Welding Stress Analysis

The weld residual stress analysis was done using commercial FEA software - ABAQUS. A special user material subroutine (UMAT) was developed to account for most of the essential features in welding process, which include the effects of stress/strain history annihilation due to material melting/remelting, weld metal deposition, and material phase changes. The details for the weld UMAT and its numerical implementation are given in literature (Ref. 3, 4). The material properties critical to weld residual stress analysis (yield strength, thermal expansion coefficient, true stress-strain curves) were determined experimentally.

1.3 Fixture FEA Structural Analysis

The objective of fixture FEA structural analysis was to evaluate the stress and distortion of the fixture due to the reaction forces when parts are pre-cambered and welded in the fixture. The analysis used the loosely coupled method. The reaction forces at various clamping locations were obtained from the welding simulation where the clamping locations were assumed to be

fixed. The forces were then applied to the fixture model and the stress and distortion were assessed using linear FEA structural analysis.

2. Case Study

In this case study, a pre-cambering robotic welding process was used to control welding distortion to eliminate post-weld treatment such as straightening and machining. Welding simulation and weld fixture stiffness analysis were the major tasks to define a pre-cambering robotic welding process that would hold the flatness tolerance after welding.

2.1 Welding experiments

A large fabricated structure, as shown in Fig. 2, was robot welded using GMAW process. The weld material is equivalent to ASTM A572 steel. There are total 26 welds in consideration in the product, as indicated by the red arrows in Fig. 2. The total welding length is about 20 meters. The welding sequence was optimized for minimum distortion based on welding simulation results. Single-pass 10 mm and 6 mm fillet welds were used on top and bottom sides, respectively. The welding parameter for 10 mm welds were: 33.7 V, 400 A, and 4.8 mm/s. The welding parameters for 6 mm welds were 23.5 V, 265 A, and 7.0 mm/s. Prior to welding, the structure was tack welded and pre-cambered in the fixture. The magnitudes and direction at various pre-cambering locations varied depending on the predetermined results from the welding simulation.

[graphic][graphic][merged small][merged small]

Fig. 2: Welds in the studied structure: (a) top side and (b) bottom side.

2.2 FEA models The FEA models for the weld product and weld fixture are shown in Fig. 3(a) and (b), respectively. The FEA model for the weld product was built with hexahedral solid elements. There were 19,820 elements and 25813 nodes in the weld product model. The size of elements within and near the welds was very fine so that welding induced thermal and mechanical features can be captured. In contrast, the size of the elements far away from weld line was very coarse for efficient computation. The weld fixture FEA model consisted of three types of elements - solid, shell, and beam elements. There were 74,374 elements and 80,260 nodes in the weld fixture model.

« PreviousContinue »