FieldsMachining and mechanical engineering
Experiments here usually have about 3 factors and 16 runs. Most are full factorials.
- Full factorial
- Fractional factorial or screening
- Central composite or Box-Behnken
- Other designs
Show the numbers
| Kind of design | Experiments | Factors | Runs |
|---|---|---|---|
| Full factorial | 21 | 3 | 16 |
| Fractional factorial or screening | 4 | 6.5 | 17 |
| Central composite or Box-Behnken | 1 | 3 | 24 |
| Other designs | 5 | 2 | 16 |
Factors and runs are the medians of each kind.
Full factorial
Every combination of the factor levels.
- Product uniformity against pressure, table speed and down force: a 2^3 run twiceNIST/SEMATECH e-Handbook of Statistical Methods (2020) · 3 factors, 16 runs
- Strength of sintered reaction-bonded silicon nitride against five machining factors: a 2^5 full factorialNIST/SEMATECH e-Handbook of Statistical Methods (2020) · 5 factors, 32 runs
- Effective life of a cutting tool at three tool angles and three cutting speeds, a 3^2 factorial run twice (coded data)Montgomery (2013) · 2 factors, 18 runs
- Life of a machine tool: cutting speed, tool geometry and cutting angle, a 2^3 run three times (Problem 6.1)Montgomery (2013) · 3 factors, 8 runs
- Life of a machine tool: replicate I of Problem 6.1 with four centre runs (Problem 6.6)Montgomery (2013) · 3 factors, 12 runs
Show the rest
- Surface finish of a metal part at three feed rates and four depths of cut, run three times (Problem 5.4)Montgomery (2013) · 2 factors, 36 runs
- Surface roughness of a machined part: tool angle, cutting fluid viscosity, feed rate and cooler, a single replicate of a 2^4 (Table P6.9) (Problem 6.35)Montgomery (2013) · 4 factors, 16 runs
- Thrust force of a drill press at two drill speeds and four feed rates, run twice (Problem 5.9)Montgomery (2013) · 2 factors, 16 runs
- Vibration of a printed circuit board cut by a router: bit size by cutting speed, a 2^2 run four times (Problem 6.5)Montgomery (2013) · 2 factors, 4 runs
- Application of coloring inks to package labels: speed, pressure and distance of a printing machine in a 3^3 factorial replicated three times (Exercise 7.9)Box and Draper (2007) · 3 factors, 27 runs
- Circular runout of a machined lathe part: finish cut, chuck jaws, clamping pressure and feed rate in a 2^4 with three parts a run (Ampex)Schmidt and Launsby (2000) · 4 factors, 16 runs
- Rivet head protrusion against countersink, the h parameter and head diameter, in four unbalanced blocks (regression on the data collected)Schmidt and Launsby (2000) · 5 factors, 27 runs
- Surface finish of a machined steel rod: RPM, feed rate and tool radius in a 2^3 factorial, four measurements a run (Problem 2 of chapter 2)Schmidt and Launsby (2000) · 3 factors, 8 runs
- Tool life against feed and speed (Problem 8 of Part IV)Box, Hunter and Hunter (1978) · 2 factors, 6 runs
- Tool life against speed and feed on a lathe, a 3^2 (Problem 29 of Part IV)Box, Hunter and Hunter (1978) · 2 factors, 9 runs
- The 2^3 run in two blocks of four, the block confounded with the three-factor interactionNIST/SEMATECH e-Handbook of Statistical Methods (2020) · 3 factors, 8 runs
- The polishing example's 2^3 in coded units and standard order, and its full modelNIST/SEMATECH e-Handbook of Statistical Methods (2020) · 3 factors, 8 runs
- The polishing example's 2^3 replicated twice, in standard order, with its settings in actual unitsNIST/SEMATECH e-Handbook of Statistical Methods (2020) · 3 factors, 16 runs
- The polishing example's randomized, replicated 2^3 with three centre runs, and its worksheet in actual unitsNIST/SEMATECH e-Handbook of Statistical Methods (2020) · 3 factors, 19 runs
- The polishing example's replicated 2^3 in a random run orderNIST/SEMATECH e-Handbook of Statistical Methods (2020) · 3 factors, 16 runs
- The polishing example's replicated 2^3 in a random run order, with three centre runsNIST/SEMATECH e-Handbook of Statistical Methods (2020) · 3 factors, 19 runs
Fractional factorial or screening
A planned part of the combinations, to find the factors that matter.
- Profile deviation of jet engine impeller blades from a five-axis CNC machine: eight machine settings in a 2^(8-3) in four blocksMontgomery (2013) · 8 factors, 32 runs
- Cutting-tool temperature: a first-order model in ln T for five variables in a 2_V^(5-1), each run read twice (Exercise 5.23, Table E5.23)Box and Draper (2007) · 5 factors, 16 runs
- Dispense volume of a chemical dispensing system: operator and seven sources of variation in a Taguchi L18, four readings a runSchmidt and Launsby (2000) · 8 factors, 18 runs
- Pressure, table speed and down force: a 2^(3-1) with D = P*T, replicated twice, with five centre runs, in run orderNIST/SEMATECH e-Handbook of Statistical Methods (2020) · 3 factors, 13 runs
Central composite or Box-Behnken
Extra runs that fit curvature, to find an optimum.
Other designs
Optimal, blocked, split-plot and other designs made for the problem.
- Assembly time in a 4 x 4 Graeco-Latin square with workplaces added (Problem 4.36)Montgomery (2013) · 4 factors, 16 runs
- Assembly time of four methods in a 4 x 4 Latin square of order of assembly and operators (Problem 4.23)Montgomery (2013) · 3 factors, 16 runs
- Shape factor of five nozzle designs at six jet efflux velocities as blocks (Problem 4.10)Montgomery (2013) · 1 factor, 30 runs
- Depth of a crater worn by a carbide cutting tool over a grid of positions, 50 observations, for second- and third-order surfaces (Exercise 11.24)Box and Draper (2007) · 2 factors, 50 runs
- Taylor's tool life equation: two variables at uncoded settings with four observations each (Exercise 7.17)Box and Draper (2007) · 2 factors, 9 runs