A sweeping plan also known as clearing designs, generally refers to experimental design methods used to explore the effects of different variables on a response. These designs are particularly useful for gaining an overview of the behavior of a system or process. The initial aim of these experimental designs is to evaluate the effects of several factors X on a response Y, and they are particularly useful in the case of a large number of factors X.
Ellistat Data Analysis offers scan plans for factors that can have between 2 and 3 levels, such as :
- L12 Orthogonal Arrays (Taguchi Designs) are part of the orthogonal planes developed by Japanese statistician Genichi Taguchi. They are widely used in experimental design to test the effects of several factors with a minimum number of experimental trials. The L12 Orthogonal Array (Taguchi Design) is a type of orthogonal plane Taguchi, which is designed to handle up to 11 factors.
- Visit experimental designs L20 is one of the orthogonal designs used for process optimization and investigating factor effects in multi-factor experiments. L20 designs allow up to 19 factors to be tested, each at two levels, while maintaining the orthogonality of the tests.
Example: Experimental design - Scanning design
ELLIMETAL manufactures complex metal parts, such as engine and transmission components, which require high precision and optimum surface finish. The milling machine used in production plays a crucial role in determining the final characteristics of the parts.
Let's imagine you're responsible for optimizing a milling machine used to manufacture metal parts. You want to determine the impact of various machine and processing parameters on the dimensional accuracy and surface finish of the parts.
The output data to be optimized is Roughness Ra (µm).
Factors to consider
The 9 factors you want to study could include:
- Spindle speed (RPM) Machine spindle speed. [1000 ; 2000]
- Tool feed (mm/min) Speed at which the tool feeds into the workpiece. [50 ; 100]
- Cutting depth (mm) Depth of cut in the material. [1 ; 2]
- Cutting tool type Tool material and shape. [A; B]
- Type of lubricant Type of lubricant used for milling. [Lx; Ly]
- Room temperature (°C) Workpiece temperature during milling. [20 ; 30]
- Cooling pressure (bar) Applied coolant pressure. [5 ; 10]
- Cutting angle (°) Angle of cutting tool in relation to workpiece surface. [15 ; 30]
- Clamping conditions : Method of mounting the part on the machine. [M1; M2]
Strategy
L12 generates 12 experimental combinations from these factors and levels, as illustrated above. Each combination represents a unique milling machine setting.
Test matrix generation with Ellistat - Scanning plan
- Click on the "DOE"then click on the map "Sweeping plan ".
- In the zone 1You can set the tab name and change the number of levels. 📝: Set the tab name to "ELLIMETAL", this will create a new tab in the grid page.
- In zone 2, we can put the number of factors, name them then put the value of the min and max 📝: Put 9 factors :
- Spindle speed (RPM) [1000 ; 2000]
- Tool feed (mm/min) [50 ; 100]
- Cutting depth (mm) [1 ; 2]
- Cutting tool type [A; B]
- Type of lubricant [Lx; Ly]
- Room temperature (°C) [20 ; 30]
- Cooling pressure (bar) [5 ; 10]
- Cutting angle (°) [15 ; 30]
- Clamping conditions [ M1; M2]
- In zone 3, you'll find a preview of the test matrix, which you can create in the create tab in zone 1. You can also create a D-optimal design, which is a type of experimental design that aims to maximize the statistical efficiency of an experiment while minimizing the number of trials required. 📝: Click on "Create DOE".
- In the "ELLIMETAL" gridIn this case, we find the test matrix with the 8 tests forming an orthogonal matrix.
- In the next column you can manually add the trial run results as shown in the following figure: DOE BALAYAGE data
