## Weight Reduction and Sensitivity Analysis in Bus Body Structure

Nowadays, there is huge competition between companies in order to make their product safer, lighter and cheaper. OEM's are continuously adopting techniques to reduce the design cycle time by reducing the number of iterations in the design phase. One such technique adopted by ASI in order meet costumer's expectation is design optimization at the concept stage. This process reduces the overall design cycle time by 30% as most of the optimizations are performed at the concept stage itself. Hence less number of iterations at the later design stages. The article contains the study of how design optimization at the initial stages is fruitful at the later stages of the design.

### Advantages of Design Optimization:

**Design Process Time Reduction: **The graph below compares the conventional and ASI design process. The pictorial comparison below clearly depicts that the extra time invested at the initial design phase leads to overall time saving of 30%. 20%-time addition in 1D optimization leads to 50 % time saving in 2D optimization design phase and 60 % time saving in CAD model update phase as the number of iterations (Design changes) at the later stages is reduced to half. The iterations are reduced as the structure is optimized at the initial stage itself hence major changes due to boundary conditions and load cases are eliminated.

The chart below shows the comparison between the 1D & 2D simulation on the final optimized model. The analysis has been done considering the correlation factor between the 1D and 2D analysis. This correlation factor was calculated using the results of different types of 1D and 2D models. Below are the results and the optimization process for the simulation activity.

**Weight Reduction:** Through structural optimization using few important load cases, the weight of the structure can be optimized. Using the same concept, a bus body structure of 2000 kg (approx.) was reduced to 1500 kg at the initial stage itself. (25% weight reduction) (Refer Figure 3 below). Process explained in detail in the Case Study Section.

The optimization at the initial stage is justified because if the structure is not optimized at the initial stage, final weight of the structure will be high due to weight addition at the later stages of product development.

**Material utilisation and complexity:** Using the above process the designers have full flexibility to reduce the type of members used in the design. For e.g., If the optimization software outputs cross sections of three members as 50x50x1.8, 48x46x1.4 and 45x47x1.9, so the designer can commonize all the three cross sections to 50x50x2. This will reduce the complexity of design and also will be favourable for manufacturing. Commonization will also reduce the complexity of supply chain process.

**Better Understanding of the Structure at the initial stages:** By sensitivity analysis, the engineers can tell which member is crucial for a particular load case and which are not. Hence, using these results the designers can focus on the critical design regions as per their load cases and boundary conditions. Better understanding of structure sensitivity at initial phase enhances flexibility in the later stages of the design.

Figure 4 below displays the stress plot for bending load case of the single bay structure of the bus. The members with high stress response are more sensitive to the particular load case than the members that are not i.e. having low stress response. After viewing these results the designer can create a list of members which he can target in order to optimize his structure for a particular load case. The same procedure can be adopted for multiple load cases.

**Methodology: **

The optimization design process work flow is given below:

1D mesh model is prepared using the initial structure line model, material data and member details. Below are the three load cases that are considered for optimization with common objective to reduce weight. The passing values for 2D CAE model are given in the bracket considering the correlation factor between the 1D and 2D model.

**1. Free- Free Modal Load Case**

Objective Function: Minimize weight

Constraint: 1st Modal Frequency (Torsion Mode) > = **5.4 Hz (5 Hz for 2D)**

Design Variables: The cross sections of the structural members.

**2. Bending Load Case**

Objective Function: Minimize weight

Constraint: Max Deformation < = **6 mm (12 mm for 2D)**

Design Variables: The cross sections of the structural members.

**3. Torsion Load Case**

Objective Function: Minimize weight

Constraint: Torsional Displacement <= **7.4 mm (9.4 mm for 2D)**

Design Variables: The cross sections of the structural members.

Full model is optimized as per the above mentioned process flow and final output is the fully optimized structure as per the load cases and boundary conditions. The correlation error between the 1D and 2D simulation has been considered in the simulation targets. The plus point about the above process is that if the structure fails in a particular load case, only one optimization cycle is required to get the optimum passing results.

**Results:**

The figure above displays the entire optimization process flow. The optimization is done as per targets of each and every load case and the model are updated as per the results. The results of the activity are the list of optimized cross sections that will meet the target values.

The table above displays the validity of the current 1D optimization process. 2D model of the final optimized structure was created and solved for all three load cases. It was found that the if we consider the correlation factor between the 1D and 2D analysis results for our optimization targets, then the actual vehicle targets can be met by this process.

In the above study, CAE optimization tool is only used for size optimization of structures using linear load cases. The scope of optimization is much diverse nowadays. The structures and components can be optimized using linear and non-linear analysis using both static and dynamic load cases (For e.g.: Optimization of a bus structure using rollover analysis (Dynamic and non-linear load case) as design control variable). The accuracy of optimization will depend upon accuracy of loads and precise assessment of results. At ASI, we have a design team that can performs all types of optimization analysis which helps the components become safe, light and cheap.

---

Pratik Shukla