Analyzing Stress Test Results for Optimal Garment Design

Part 1: Reading the Visual Language of a Stress Test

Before we dive into the statistical analysis of a full factorial experiment, let's start with the most fundamental piece of data from a single tensile test: the load-deformation curve. This graph is a direct visual representation of how a fabric behaves under stress.

Look closely at the image above. It shows the results of a uniaxial tensile test on a sample of knit fabric.

• The Y-axis (Load): Represents the force being applied to the fabric.
• The X-axis (Change in length): Represents how much the fabric stretches (strain) in response to that force.

The shape of the curve tells a story. In the "Wale direction" graph, you can observe several key behaviors:

1. Initial Resistance (Origin to point c): The curve starts with a certain slope. A steeper slope indicates a stiffer material that resists stretching (high modulus).
2. Yielding/Deformation: As the load increases, the loops of the knit structure begin to deform and straighten out.
3. Jamming (point d onwards): The slope of the curve increases sharply. This phenomenon, known as "jamming," occurs when the knit loops have deformed to the point where they are pressed tightly against each other. At this stage, the fabric behaves more like a solid sheet, and it takes a much larger force to produce any further stretch.

Notice the difference between the course and wale direction graphs. The fabric is significantly stiffer and stronger in the wale direction. This anisotropy is a critical property of most textiles and underscores why biaxial testing, which we discussed previously, is so important for understanding performance in a garment that is stretched in multiple directions simultaneously.

Part 2: Statistical Interpretation of a Designed Experiment

Now, let's apply your statistical expertise to interpret the results of the 2^4 factorial experiment on seam strength we designed in the last lesson. We have hypothetically run the 16 tests, and now we have the data. Our goal is to determine which factors—Stitch Density, Thread Type, Seam Type, Needle Size—have a significant impact on our response variable, Seam Breaking Strength.

The research paper "Determination of Key Characteristics of Seam Strength" provides a perfect template for this analysis.

Let's break down how you, as the designer-engineer, would use this information.

Step 1: Identify What Matters with an Effects Table

As you saw in Table 5 of the paper, each factor's influence can be quantified. The "Percent Contribution" column is particularly illuminating. In that study, the most influential factors were:

• Factor D (Stitch density): 81.2%
• Factor E (Fabric tensile strength): 8.7%
• Factor C (Fabric stability): 3.3%

All other factors and interactions contributed less than 2% each. This is a critical finding. It tells you that if you want to engineer a strong seam, you must pay the most attention to stitch density and the properties of the base fabric itself. The effect of, for example, sewing thread tension (Factor A, 0.23%) is negligible in comparison. This allows you to focus your efforts where they will have the most impact.

Step 2: Determine the Optimal Direction with Main Effect Plots

Knowing that a factor is significant isn't enough; you need to know which level is better. This is where main effect plots, like those in Figure 3 of the paper, are invaluable.

A main effect plot shows the average response value at each level of a factor.

• If the line goes down from the low level (-1) to the high level (+1), it means the response is better (in this case, higher seam strength, though the paper's graph is counter-intuitive) at the low level.
• If the line goes up, the high level is better.
• If the line is nearly flat, the factor has little to no effect.

The paper concludes that for factors C, D, and E, the lower level produced higher seam strength. For example, for Stitch Density (D), the lower level (8 SPI in their experiment) was better than the higher level (12 SPI). This is a direct, actionable instruction for construction.

Step 3: Check for Complex Relationships with Interaction Plots

Sometimes the effect of one factor depends on the level of another. This is an interaction. As you know from DoE, interaction plots are the key to spotting these. Non-parallel lines indicate a potential interaction. In the paper, the CD interaction was marginally significant. This suggests that the optimal stitch density might change depending on the fabric's stability. While this effect was small in their study, in your own experiments for a Bollywood costume, you might find a strong interaction between, for instance, fabric type and seam type. This would mean a flatlock seam might be strongest on powernet, while a superimposed seam is strongest on tricot. Ignoring such an interaction would lead to suboptimal design choices.

Part 3: From Analysis to Selection

You have now analyzed the data and identified the statistically significant factors and their optimal levels. The final step is to translate this statistical conclusion into a practical design decision, integrating it with a broader understanding of material science.

Case Study: Selecting a Fabric for a High-Impact Bandeau Top

• Scenario: You are designing a bandeau bikini top for a dance number that involves high-energy, athletic movements. The bandeau must provide excellent support and, crucially, must not stretch out or "grow" during the performance, which would lead to a catastrophic failure of support.
• Experiment: You've run a 2^2 factorial experiment testing two factors:
  1. Fabric: Standard powernet vs. High-performance powernet.
  2. Seam Finish: Standard serged seam vs. Seam reinforced with stabilizing tape.
• Response Variable: You didn't measure breaking strength. Instead, you measured Permanent Strain (%) after a 30-minute simulated high-impact biaxial test. A lower value is better.

Interpreting Your Hypothetical Results:

Your analysis shows a large and significant main effect for Fabric and a smaller, but still significant, main effect for Seam Finish. The interaction is not significant. The main effect plot shows that High-performance powernet resulted in much lower permanent strain.

How do you translate this? This is where we connect the statistics back to material properties.

Your experimental result—that High-performance powernet had less permanent strain—can now be explained in material science terms. The fabric likely has a higher modulus (it's more resistant to being stretched in the first place) and a superior stretch recovery rate (it snaps back to its original shape more completely). The statistical result validates the choice of the more expensive, higher-performance material.

This FAST chart shows how these properties can be benchmarked. For your high-impact bandeau, your analysis tells you to select a fabric that falls into the "STIFF" range for Bending Rigidity and has low "Extensibility" (e.g., 1-3%) to ensure a firm, anchoring band.

The Final Decision:

Based on your interpretation, you make the following selection:

"The optimal combination for this high-impact bandeau is the High-performance powernet constructed with a reinforced seam. The experimental data clearly shows that the fabric choice is the most critical factor in preventing permanent deformation ('growth'). This fabric's superior performance is attributed to its high modulus and stretch recovery. While the reinforced seam offers a secondary improvement, the primary driver of performance is the material itself."

You have successfully used a simulated stress test to make an evidence-based design choice.

Conclusion

Today's lesson has closed the loop on our stress-testing module. By building upon the experimental design from our previous session, you have learned to interpret both graphical and statistical outputs to make informed, optimal decisions. This process elevates design from an art based on intuition to an engineering discipline grounded in empirical data.

Key Takeaways:

• Load-deformation curves provide a direct visual summary of a fabric's stiffness, elongation, and failure points.
• Factorial experiment analysis, using tools like effect plots and ANOVA that are familiar to you, allows for the efficient identification of the most critical factors influencing garment performance.
• Main effect plots show which level of a factor is optimal, while interaction plots reveal complex relationships where the best choice for one factor depends on another.
• The ultimate goal is to connect statistical significance to physical properties. A significant result for "Fabric Type" should be linked to measurable characteristics like modulus and stretch recovery to make a defensible design selection.

Next Lesson Preview

You have now mastered the engineering principles of material selection. You can choose the right fabric and seam based on data. But a garment is more than just its materials; it's a three-dimensional object that must conform to the human body. In our next lesson, "Construct a 'toile' (muslin prototype) of a complex bra design," we will begin the process of turning your 2D pattern into a 3D reality, allowing for the first crucial assessment of structure and fit.