Key Formulas
Heat input: $Q = V \cdot I \cdot \eta / v$
Angular distortion: $\theta = C_1 Q / t^2$
Longitudinal shrinkage: $\delta_L = C_2 Q / (EA)$
Peak residual stress ≈ yield stress $f_y$ (fully restrained)
Adjust heat input, joint type, thickness, material, and restraint to instantly compute residual stress distribution, angular distortion, and longitudinal shrinkage. PWHT stress-relief effects included.
Heat input: $Q = V \cdot I \cdot \eta / v$
Angular distortion: $\theta = C_1 Q / t^2$
Longitudinal shrinkage: $\delta_L = C_2 Q / (EA)$
Peak residual stress ≈ yield stress $f_y$ (fully restrained)
The core driver of all welding effects is the heat input per unit length, which determines how much metal melts and subsequently contracts.
$$Q = \frac{V \cdot I \cdot \eta}{v}$$Where $Q$ is the heat input (J/mm), $V$ is voltage (V), $I$ is current (A), $\eta$ is the arc thermal efficiency, and $v$ is the travel speed (mm/s). This is the primary parameter you control in the simulator. Higher $Q$ means more heat, leading to greater shrinkage and stress.
The resulting deformations are often empirically related to heat input and geometry. Angular distortion (bending) depends heavily on plate thickness, while longitudinal shrinkage depends on the cross-sectional area resisting the contraction force.
$$\theta = C_1 \frac{Q}{t^2}\quad \text{and}\quad \delta_L = C_2 \frac{Q}{E A}$$Here, $\theta$ is the angular distortion (radians), $t$ is plate thickness (mm), $\delta_L$ is longitudinal shrinkage (mm), $E$ is Young's modulus, $A$ is the weld cross-section area, and $C_1, C_2$ are empirical constants. The simulator solves these instantly when you adjust thickness or joint type.
Shipbuilding & Offshore Structures: Massive welded panels are prone to buckling distortion (a type of angular distortion). Engineers use simulators like this to optimize welding sequence and heat input before production, saving millions in rework. Preheating thick high-strength steel joints is standard to prevent cracking.
Pipeline Construction: Longitudinal shrinkage in long pipeline welds can accumulate, causing the pipeline to shorten beyond allowable tolerances. The travel speed and heat input are carefully controlled based on calculations similar to this tool's output to ensure fit-up for the next pipe section.
Pressure Vessel Fabrication: High residual stress at welds can combine with internal pressure stress, leading to premature failure or reduced fatigue life. Post-Weld Heat Treatment (PWHT), which you can model with the "PWHT hold time" parameter, is often mandated to relax these stresses and ensure safe operation.
Automotive Frame Assembly: Distortion in chassis components from welding can misalign mounting points for engines or suspensions. CAE engineers use this physics to simulate welding during virtual prototyping, adjusting clamp positions (modeled by "Restraint Condition") to minimize warpage before building physical tools.
First and foremost, keep in mind that this simulation is not an "all-knowing oracle." If you put in arbitrary input parameters, the output will be "garbage in, garbage out." For example, material property data. Many people use catalog values for room temperature data as-is, but in welding simulations, properties like Young's modulus and yield stress at 600°C or 800°C significantly influence the results. The room temperature Young's modulus for SUS304 is about 193 GPa, but at 800°C, it drops to less than half, around 90 GPa. If you get this data wrong, both the deformation and residual stress results will be way off the mark.
Next, errors in constraint condition settings. In reality, the workpiece is firmly fixed by jigs, but if you apply a fully fixed condition (constraining all degrees of freedom) in the simulation, it can sometimes calculate unnaturally high residual stresses. Conversely, if the constraints are too loose, the deformation may appear larger than in reality. For instance, for long butt joints, settings that imagine the actual jig—like constraining only certain directions to allow for thermal expansion-induced "snaking"—are essential.
Finally, don't overlook mesh dependency. Sharp temperature gradients occur around the weld bead, so if you don't use a fine mesh there, you won't accurately capture the temperature or stress fields. However, making the entire mesh too fine will cause computation time to explode. A standard practice is "graded mesh refinement"; for example, using a 1mm mesh near the weld line and a 5mm mesh farther away for a 20mm thick joint. If doubling the mesh density doesn't significantly change the results, you can be reasonably confident.
The "residual stress" and "deformation" you calculate with this tool aren't just about welding. They are directly connected to fatigue strength analysis. For example, if a tensile residual stress exists at the weld of an automotive suspension arm, it combines with the cyclic loads (mean stress) during driving, drastically reducing the fatigue life. By importing the residual stress distribution from the simulation as an initial condition for fatigue analysis in another CAE software, you can achieve a more realistic life prediction.
Another field is fracture mechanics. In components like power plant piping, tiny cracks can exist at welds. The "crack driving force" that evaluates whether such a crack will propagate is heavily influenced by the surrounding residual stress field. A compressive residual stress makes it harder for the crack to open, while a tensile stress can lead to rapid propagation. The stress distribution revealed by NovaSolver becomes a crucial input for fitness-for-service assessments based on fracture mechanics.
Furthermore, the influence of materials engineering, particularly phase transformation, cannot be ignored. When welding high-strength steel or carbon steel, rapid cooling can form hard and brittle martensitic structures. The volumetric expansion accompanying this phase transformation creates complex residual stress patterns that cannot be captured by simple thermal expansion/contraction calculations. In advanced simulations, you might even couple with a "metallurgical model" that predicts microstructural changes from the thermal history.
As a recommended first step, learn the basic theory of thermo-elasto-plastic analysis. Tools like NovaSolver perform calculations based on this theory behind the scenes. The key concepts are "thermal strain" and "plastic strain." When a material is heated, it strains due to thermal expansion (thermal strain). If this is constrained by the surroundings, plastic deformation (permanent, non-recoverable deformation) occurs, leaving behind forces after cooling (residual stress). Try to understand this sequence while visualizing a stress-strain diagram.
Regarding the mathematical background, understanding the basics of the Finite Element Method (FEM) and how the system of equations is solved (solver technology) will help you grasp the reliability of results and the meaning of error messages. A key point is understanding how the heat conduction analysis (temperature field calculation) and structural analysis (stress field calculation) are solved sequentially or in a coupled manner (sequentially coupled analysis). Most of the time-consuming computation is spent converging this nonlinear problem.
If you want to delve deeper for practical work, look into experimental validation methods. How do you verify simulation results? For example, residual stress is measured using "X-ray diffraction" or "neutron diffraction," and deformation is measured with a "3D scanner." The experience of comparing and calibrating your simulation model against experimental data from simple test specimens is arguably the best learning experience, dramatically improving your model setup accuracy.