Leveraging simulation to make more efficient use of packaging material and reducing costly physical testing

Customer profile

Haier Air Conditioner (Haier) is one of the world’s largest manufacturers of home appliances and consumer electronics. The company is the leader of its industry in China, where it is headquartered.

Haier ships its products across the globe and in more than 100 countries, so well-designed product packaging is crucial to the company’s ability to deliver products without damage to even the most distant destinations.

The challenge: a lower-cost way to prevent drop damage

While the company is known for its quality products, the air conditioners it manufactured were being damaged during transportation, leading to higher costs and delays in final product delivery.

Haier conducted physical drop tests to try to improve the structure of its air conditioners as well as its packaging to make them more resistant to drop damage. These tests, however, significantly increased the products’ research-and-development (R&D) costs, and they consumed an extraordinary amount of time.

In physical tests, engineers could not easily observe the damage process because the collision between the product and the ground was an instantaneous event. They could view the outcome but not the strains and shape changes during the fractions of seconds in which they occurred.

“The deformation time in the drop test is typically less than 50 milliseconds,” says Yundong Chen, a Haier Engineer. “The model is packed in a nontransparent box, which makes it difficult to observe the internal deformation process.” As a result, Haier considered using excessive packaging materials, but the overall design strength of the package was insufficient.

The solution: using Simcenter Hypermesh to simulate drop testing

Haier engineers turned to Simcenter™ software from Siemens to create a virtual simulation of a drop test that would provide extensive data on the stresses and strains involved. To conduct virtual drop tests, they used Simcenter Hypermesh™ software for preprocessing and the Simcenter Radioss® software solver for two types of air conditioners to gain better insights.

For the first product, Haier engineers initially conducted numerous physical tests to collect data on the performance of the materials used in the packaging, such as honeycomb panels and EPS blocks. Then it simulated material tests with virtual models to calibrate the data.

They tested honeycomb paper with elastic deformation, buckling, plastic collapse and brittle rupture of the product under axial compression. After obtaining the strain-stress curve from the physical model, the team simulated the honeycomb paper and reproduced the axial compression virtually. The virtual compression test curve showed very good conformity with the experimental physical test curve.

The engineers employed Simcenter Hypermesh in the preprocessing of both the air conditioner structure and its packaging. To complete the simulation, they set an initial velocity for a drop and added gravity as a factor.

They used the Simcenter Radioss explicit solver to perform the drop-test analysis, and Simcenter Hyperview™ software and Simcenter Hypergraph™ software to generate reports on the results. Simcenter enabled Haier to carry out the entire simulation process within a single computer-aided engineering (CAE) environment.

For the second type of air conditioner, Haier engineers again prepared separate models for the air conditioner and the packaging and then employed Simcenter Radioss in simulating a 0.8-meter drop on a corner of the package, with a beginning speed of 3.96 millimeters (mm)/millisecond (ms). Engineers subsequently conducted follow-up simulations for other conditions.

“By using Simcenter Radioss for a fast and accurate solution,” says Chen, “we could study all the parameters, such as stress, strain and displacement, rather than merely looking at the air conditioner damage after real drop tests and imaging of the drop process.”

Left: Drop test of first AC unit with Simcenter Radioss.
Middle: Simulation unveils weak metal chassis.
Right: Improved bead design with Simcenter Optistruct topography optimiz.

The results: drop damage eliminated

For the first air conditioner, the initial drop simulations showed a substantial deformation when the product was dropped on one corner.

Engineers could see from the simulation that the strength of the corners of the machine itself was not sufficient to withstand such a drop, with the local plastic strain reaching 40.2 percent. Additionally, engineers determined the thickness of the packaging pad was too small to serve as a buffer during a drop, allowing the product’s chassis to recess when it struck the ground.

Haier redesigned the structural base of the air conditioner as well as the packaging and then conducted the same drop-test simulation with a refined model.

“This test revealed that the structural strength of the machine corner had been strengthened in the refined model,” Chen says, “with the local plastic strain dropping to 0 percent and no permanent deformation of the product.”

For the second air conditioner, conducting finite element analysis (FEA) determined the base of the appliance sustained a strain value of 71.6 percent and readily showed the problem resulted from the initial thickness and shape of the packaging design.

Engineers used Simcenter Hypermesh to quickly optimize the packaging, remodeling thickness and shape, thereby reducing the strain value to just 6.1 percent.

“This meant the new packaging design presented close to no risk of damage during a fall,” Chen observes. “We could meet all the design requirements while greatly shortening the development time. A subsequent bench test also validated the CAE findings.”

Second AC unit drop test.

Improving product design with optimization

Once Haier engineers had resolved drop-test issues, they took additional steps to optimize the design of the air conditioner chassis itself with Siemens optimization tool Simcenter Optistruct® software.

“We understood the value of looking into a solution to evaluate product performance during the development stage to shorten development time and save cost,” says Chen. “By using Simcenter Optistruct optimization techniques in the air conditioner’s structure, we felt we could increase product reliability.”

Previous methods of evaluating the mechanical structure’s stress and strain values with applied material mechanics and elasticity calculations were complicated because of complex geometry and load. Mathematical solutions often were incomplete or yielded results that were very different from the actual situation.

“While checking the strength of the mechanical structure for reliability, the selected safety factors are often too big, resulting in the size and weight of the structure design being too large,” Chen explains. “Moreover, since the calculation and analysis are rough, some weaknesses may occur.”

Left: Comparison of deformations on the original and optimized design of the AC unit foot.
Right: Improved EPS block design performs better over the original design despite 28 % in weight savings.

The results: a stronger structure with no change in composition

Haier performed a drop analysis of the chassis model using Simcenter Radioss and found the design of the chassis area that supports the compressor was simple and weak. Corners were drastically deformed in the simulations and compressor bolts appeared to loosen.

Using Simcenter Optistruct for topology optimization, engineers were able to improve the structural strength of the air conditioner chassis by 40 percent without changing its composition and test the optimized structure afterward.

Chen said that Haier engineers identified several benefits in using Simcenter tools: The reasons for parts failure could be found quickly and accurately. With Simcenter postprocessing software, engineers could see every moment of the drop testing process and extract such numerical data as stress, strain and displacement to analyze the product’s behavior.

Haier was able to use packaging and structural material more reasonably after the simulation. Once the reason for the failure was found, engineers could redesign the product in the appropriate way, reducing trial-and-error costs.

”Using Simcenter for simulation brought great efficiencies and savings to the redesign process,” says Yundong Chen, senior CAE engineer, Haier Air Conditioner. “It helped reduce the huge amount of work involved in physical tests of the product and its packaging, making redesign more efficient and eliminating the costs of physical testing.”

Thanks to the capabilities of the Simcenter Radioss solver, the simulation was highly accurate. With the multi central processing unit (CPU) technology and quality scaling capability of Simcenter Radioss, the drop-test analysis produced significant computational efficiency.