Title: Simulates LOX Drop Testing
Version: 0.1.3
Author: Chad Ross [aut, cre]
Maintainer: Chad Ross <chad.ross@gmail.com>
Description: Generates simulated data representing the LOX drop testing process (also known as impact testing). A simulated process allows for accelerated study of test behavior. Functions are provided to simulate trials, test series, and groups of test series. Functions for creating plots specific to this process are also included. Test attributes and criteria can be set arbitrarily. This work is not endorsed by or affiliated with NASA. See "ASTM G86-17, Standard Test Method for Determining Ignition Sensitivity of Materials to Mechanical Impact in Ambient Liquid Oxygen and Pressurized Liquid and Gaseous Oxygen Environments" <doi:10.1520/G0086-17>.
Depends: R (≥ 3.4.0)
License: MIT + file LICENSE
Encoding: UTF-8
LazyData: true
Imports: data.table
URL: https://github.com/chadr/droptest
RoxygenNote: 6.0.1
NeedsCompilation: no
Packaged: 2018-08-23 14:32:30 UTC; Chad
Repository: CRAN
Date/Publication: 2018-09-19 19:00:03 UTC

droptest: Simulates LOX drop testing.

Description

Generates simulated data representing the LOX drop testing process (also known as impact testing). Drop testing is expensive, time consuming, and notoriously difficult to analyze. A simulated process allows for accelerated study of test behavior.

Details

Functions are provided to simulate trials, test series, and groups of test series. Functions for creating plots specific to this process are also included. Test attributes and criteria can be set arbitrarily. This work is not endorsed by or affiliated with NASA.

Drop testing – sometimes called impact testing – is used to evaluate if a material will interact with liquid oxygen (LOX). The material is exposed to the LOX and an impactor is dropped onto the sample. Each drop is a bernoulli trial where a reaction is a failure and a non-reaction is a success. The specified number of trials – until failure – completes one test.

While fundamentally a binomial process, drop testing – performed by the military and NASA – yields results that are difficult to analyze. Numerous tech briefs and standards have attempted to address the problem (see below for more information). Testing stops immediately once the failure condition is reached. If the failure condition occurs on drop one or two – depending on the failure criteria – then the test returns only one or two result values. Alternatively, if a material passes, or if the failure condition occurs on the last trial, then the test returns as many result values as trials.

Simulation can be used to examine the behavior of this test procedure.

Inspired by NASA Technical Note "Computer Simulation of Threshold Sensitivity Determinations" (NASA-TN-D-7663). Gayle (1974). https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19750004618.pdf

Definitions

Author(s)

Chad Ross chad.ross@gmail.com

References

Pass/Fail criteria and number of observations required have been defined in the following standards:

Note: This package is not constrained by any standard. Arbitrary test criteria and observations can be specified for maximum flexibility.

For more information on drop testing:

Note: This work is not endorsed by or affiliated with NASA. Released under MIT license.

See Also

dtrials dseries dgroups trialdev gayleplot ratioplot D7905 PREGENTD

Historical Data

Description

Historical drop test data where testers did not immediately terminate testing. Useful for comparing modern test methods and simulation output.

Usage

data("D7905")

Format

A data frame with 10 observations on the following 2 variables.

P

a numeric vector

SD

a numeric vector

Source

NASA Technical Note NASA-TN D-7905. (1970) Written by J. B. Gayle. https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19750014413.pdf

Examples

data(D7905)

Trial Deviation for Pre-Generated Data

Description

Output of trialdev() on a large pre-generated dataset.

Original dataset represents fifty levels of q where each level has 100,000 simulated test series. For a total of 5,000,000 simulated tests.

Usage

data("PREGENTD")

Format

A data frame with 50 observations of 4 variables.

See Also

dseries dgroups

Examples

data(PREGENTD)

Generates groups of simulated test series.

Description

dgroups returns a collection test series organized into groups. Each batch of test series are identified with a group number. Test parameters will be uniform within each group.

Usage

dgroups(num.groups = NULL, probs = NULL, ...)

Arguments

num.groups

Integer. Specifies how many groups of drop tests to simulate.

probs

Vector. Specifies probabilities of q used for each group. Vector length must equal value of num.groups.

...

Passes values to dseries.

Value

Data table of groups where each group consists of multiple drop tests. Consisting of the following elements:

Author(s)

Chad Ross chad.ross@gmail.com

See Also

dseries dtrials droptest

Examples

  dgroups(num.groups = 2, probs = c(0.01, 0.2), num.series = 5)
  dgroups(num.groups = 2, probs = c(0.01, 0.2), num.series = 5, max.trials = 60)
  dgroups(num.groups = 2, probs = c(0.01, 0.2), num.series = 5, fail.criteria = 2)
  dgroups(num.groups = 5, probs = seq(0.01, 0.05, by = 0.01), num.series = 2)

Completes multiple series of simulated drop tests.

Description

dseries returns a series of simulated drop tests. Where each test consists of trials (drops).

Usage

dseries(num.series, tag.group = FALSE, group = 0, ...)

Arguments

num.series

Integer. Specifies how many series of drop tests to simulate.

tag.group

Logical. If TRUE, group number is included in output data table. Default is FALSE.

group

Integer. Assigns a group number to each simulated drop test within the same series. Only if tag.group is TRUE. Default is 0.

...

Passes q, max.trials, fail.criteria, and fail.criteria to dtrials. All are optional except q.

Value

Data table of multiple simulated drop tests. Each row of the data.table represents one simulated drop test. Containing the following elements:

Author(s)

Chad Ross chad.ross@gmail.com

See Also

dtrials droptest

Examples

  dseries(num.series = 5, q = 0.05)
  dseries(num.series = 5, q = 0.05, max.trials = 60)
  dseries(num.series = 5, q = 0.05, max.trials = 60, fail.criteria = 2)

Completes a simulated drop test.

Description

dtrials returns simulated bernoulli trials (drops) that compose one drop test.

Usage

dtrials(q, max.trials = 20, fail.criteria = 1,
  data.structure = "data.table")

Arguments

q

Integer. Specifies the probability that a reaction occurs. A reaction is interpreted as a failure therfore q is the probability of failure for the bernoulli trials. Probability of success would be p = 1 - q. See https://en.wikipedia.org/wiki/Bernoulli_trial

max.trials

Integer. The maximum number of bernoulli trials to perform. where each trial represents one drop of the impactor onto a sample. Trials performed will always be less than or equal to max.trials. Default is 20.

fail.criteria

Integer. Specifies number of reactions (failures) that can occur before an entire test is considered a failure. Must be less than or equal to max.trials. Default is 1.

data.structure

Instructs function to return result as a data.table or list. Default is data.table.

Note: Other functions in this package only work with data.tables. List is an option strictly for future flexibility.

Value

A data table or list. Containing the following elements:

Author(s)

Chad Ross chad.ross@gmail.com

See Also

droptest

Examples

  dtrials(0.05)
  dtrials(0.05, max.trials = 60)
  dtrials(0.05, fail.criteria = 2)
  dtrials(0.05, max.trials = 60, fail.criteria = 2)
  dtrials(0.05, data.structure = "list")

Plot of std deviation percent, vs probability of reaction percent.

Description

gayleplot Creates plot similar to the one seen in NASA Technical Note NASA-TN-D-7905. (1970) Written by J. B. Gayle. The namesake of this function. https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19750014413.pdf

Usage

gayleplot(simq = seq(0.01, 0.5, by = 0.05), ...)

Arguments

simq

Vector. Probabilities of failure (q) to use for simulated trials. Defaults to range of q = 0.01 to q = 0.5 in 0.05 steps.

...

Passes values to dgroups.

Details

Plots historical standard deviations ( mid-test – against binomial process. Also plots standard deviations ( from simulation using modern procedure.

Shows how historical data follows a binomial process, but simulated data produced using modern procedure does not.

Value

none

Author(s)

Chad Ross chad.ross@gmail.com

See Also

dgroups dseries dtrials droptest

Examples

  # recommended to use num.series value greater than or equal to 1000.
  gayleplot(num.series = 100)
  gayleplot(num.series = 100, simq = seq(0.01, 0.05, by = 0.01))

Creates plot of pass/fail ratio for repeated test series.

Description

ratioplot Creates plot of pass/fail ratio for repeated test series. By simulated probability of reaction.

Usage

ratioplot(simq = seq(0.01, 0.2, by = 0.01), colors = c("RED", "GREEN"), ...)

Arguments

simq

Vector. Probabilities of failure (q) to use for simulated trials. Defaults to range of q = 0.01 to q = 0.5 in 0.05 steps.

colors

Vector. Passes color options to barplot. Defaults to red and green.

...

Passes values to dgroups.

Value

none

Author(s)

Chad Ross chad.ross@gmail.com

See Also

dgroups dseries dtrials droptest

Examples

 # recommended to use num.series value greater than or equal to 1000.
 ratioplot(simq = seq(0.01, 0.20, by = 0.01), num.series = 100)
 ratioplot(simq = seq(0.01, 0.20, by = 0.01), num.series = 100,
           colors = c("BLUE", "BLACK"))

Calculates trial deviation for simulated trials (drops).

Description

trialdev creates a data.table with the average distances from q for the total percent of reactions (failures).

Usage

trialdev(sim.values)

Arguments

sim.values

Data table. Data table produced by droptest::dtrials, droptest::dseries, or droptest::dgroups.

Value

Data table of q, p, trial deviation, and average total trials per test. Aggregated by q.

Author(s)

Chad Ross chad.ross@gmail.com

See Also

dtrials dseries dgroups droptest

Examples

  trialdev(dtrials(q = 0.05, max.trials = 60, fail.criteria = 2))