DoE Design

DoE Design plans a series of test prints. You name the settings that should be tuned, give the lowest and highest value for each, and choose a plan type. The component produces a run table with one row per test print and one column per setting, plus empty columns for the measurements. The table can be exported to Excel or CSV and later read by RSM Evaluate.

Panel: 6 Calibrate. Nickname: DoE. Compute button: no (runs on every change).

Inputs

Name Type Default Description
names Text (list) layer_height; print_speed; pressure Factor names, one per item, for example layer_height, print_speed, pressure
lows Number (list) 2.5; 30; 1.0 Factor low levels in actual units, in the same order and count as names
highs Number (list) 4.5; 50; 1.3 Factor high levels in actual units, above lows and in the same order and count as names
design Text BBD Design type: DSD (screening, 4 to 12 factors), BBD (RSM, 3 to 7), Taguchi (2 to 13), FullFact2 (1 to 10), FracFact2 (fractional two level, 4 to 9)
responses Text (list) layer_width; surface_defects; print_time Response names. They become empty columns in the exported plan for lab measurements

Optional group: Advanced options (centres, center points, levels, seed)

Right-click the component to switch this group on. The inputs appear only while the group is enabled.

Name Type Default Description
centres Number (list)   Optional center level per factor away from the midpoint, for example pressure 1.15 in [1.0, 1.3]
centerPts Integer -1 Extra or total center points (-1 = design default, BBD 3 to 6, DSD 0)
levels Integer 3 Taguchi factor levels, 2 or 3. DSD and BBD always use 3
seed Integer 0 Seed for the run order. 0 or higher randomizes with that seed, -1 keeps standard order
fracRuns Integer -1 FracFact2: target runs 8 or 16 (-1 = auto: smallest with resolution ≥ IV)
replicates Integer 1 Copies of every design row, including center points
augment Text none none keeps the plan, ccd-face appends face centered star and center runs to a FullFact2 or FracFact2 plan

Optional group: Power analysis (can this plan detect the effect?)

Right-click the component to switch this group on. The inputs appear only while the group is enabled.

Name Type Default Description
powerSigma Number 0 Expected measurement scatter in response units, greater than 0 and required with powerEffect for the power estimate
powerEffect Number (list)   Smallest response change to detect in response units, one value for all factors or one per factor
powerModel Text Linear Model that will be fitted to the data: Linear, 2FI or Quadratic. Power depends on its residual degrees of freedom
powerSims Integer 400 Number of simulated experiments for the power estimate (more gives a steadier estimate)

Optional group: Excel / CSV export inputs

Right-click the component to switch this group on. The inputs appear only while the group is enabled.

Name Type Description
xlsxPath Text Local .xlsx file or folder to export on solve with Export enabled, overwriting the target
csvPath Text Local .csv file or folder to export on solve with Export enabled, overwriting the target

Outputs

Name Type Description
runs Number (tree) Run plan in actual units, one branch per run, factor values in input order
coded Number (tree) Run plan in coded units, -1 to +1
factorNames Text (list) Factor names. Wire into the RSM component
runLabels Text (list) Run labels (BBD01, …)
table Text Readable run table. Connect a panel
info Text Design summary and warnings
power Text Monte Carlo power per factor at α = 0.05. Enable the power group and connect a panel

Back to Components.