ASTM E29 Standard

Overview and Key Concepts

By Nikhil A
Assistant Testing Engineer
Aries Oil Field Services L.L.C
Introduction

 This presentation covers the ASTM E29 standard which describes

methods for using significant digits in test data to determine
conformance with specifications.
 This standard ensures that the significant digits in test results are used
correctly to reflect the true precision of the measurements.
 To facilitate consistent and reliable decision-making regarding
compliance with specified limits and tolerances.
 Reduces ambiguity in interpreting test results, leading to better quality
control and assurance.
Significance and Use
 The standard describes two commonly accepted methods of rounding
data: Absolute Method and Rounding Method.
 The standard aids in making precise and accurate decisions regarding
product quality and compliance with specified limits and tolerances.
This is crucial for maintaining high standards in manufacturing and
testing processes.
 Absolute Method: In certain fields, all digits in an observed value or
calculated value are considered significant for determining
conformance with specifications. This method ensures strict adherence
to specified limits without rounding
 Rounding Method: In other fields, the observed or calculated values
are rounded to the nearest specified unit before comparison with the
specification limits. This method simplifies data handling and
reporting.
Significant Digits in Test Data

 Significant digits in a number are the digits that carry meaningful

contributions to its precision. They include all non-zero digits, any
zeros between significant digits, and trailing zeros in the decimal part.
 The number of significant digits should reflect the precision of the
measurement method and instrument.
Rules of Significant Digits

 Non-zero digits are always significant.

Example: 123.45 (5 significant digits)
 Zeros between significant digits are significant.
Example: 1002 (4 significant digits)
 Leading zeros are not significant.
Example: 0.0025 (2 significant digits)
 Trailing zeros in the decimal part are significant.
Example: 2.500 (4 significant digits)
 Trailing zeros in a whole number without a decimal point are
not significant.
Example: 1500 (2 significant digits)
Absolute Method

 Definition: Direct comparison of test data with absolute limits.(no

rounding)
 Application: Every digit in observed/calculated value is significant.
 Example: Spec limit = 10.0, Observed = 10.05, Result = Non-
conformance.
 Importance: Ensures precision without rounding off; all digits
matter.
Absolute Method Applications

 In the Absolute Method, an observed value or calculated value is

compared directly. Value is not to be rounded.
 Pharmaceuticals: Guarantees precise active ingredient dosages.
 Material Science: Maintains strict thickness tolerances for coatings.
 Metrology: Provides precise instrument calibration.
 Construction: Delivers exact component dimensions for perfect fit.
 Note: All values are considered significant, ensuring highest accuracy
with the specified limit without rounding.
Expression of Absolute Method
 If apply to all specified limits in the standard –”For purposes of
determining conformance with these specification, all specified limits in this
standard are absolute limits, as defined in ASTM Practice E29, for using
Significant Digits in Test Data to Determine Conformance with specifications.”
 If apply to all specified limits of some general type in the
standard (such as dimensional tolerance limits) – “For purposes
of determining conformance with these specifications, all specified
(dimensional tolerance) limits are absolute limits, as defined in ASTM Practice
E29, for Using Significant Digits in Test Data to Determine Conformance with
Specifications.”
 If apply to all specified limits given in a table - “Tolerance
limits specified are absolute limits as defined in Practice E29, for
Using Significant Digits in Test Data to Determine Conformance
with Specifications.”
Rounding Method

 Definition: Rounding test data to nearest specified limit.

 Rule: “Five-even” rule - round 0.5 towards the nearest even number.
 Example: 76.50 becomes 76, 89.50 becomes 90.
 Purpose: Minimizes rounding bias, ensuring fair and consistent
results.
 This method is used when the specification or standard directly
references the practice of rounding. It’s important to specify which
method to use, absolute or rounding, as it affects the interpretation of
the test results
Rounding Rules

 If next digit beyond retention is < 5, retain last digit unchanged.

 If next digit beyond retention is > 5, increase last digit by 1
 If next digit beyond retention is 5, and:
• No digits beyond 5: Increase by 1 if last digit is odd; else, leave
unchanged.
• Non-zero digits beyond 5: Increase last digit by 1.
Examples of Rounding Method

 Digits Less Than 5:

Rule: Retain the digit in the last place.
Example: 6.423 rounded to two decimal places is 6.42.

 Digits Greater Than 5:

Rule: Increase by 1 the digit in the last place retained.
Example: 6.427 rounded to two decimal places is 6.43.
Example of Rounding Method(contn)

 Digits Equal to 5:
Rule: If there are no digits beyond 5 or only zeros, increase by 1 if the last
retained digit is odd, leave unchanged if even.Increase by 1 if there are
non-zero digits beyond 5.
Example: 6.425 rounded to two decimal places is 6.42 (last retained digit
even), 6.435 rounded to two decimal places is 6.44 (last retained digit
odd).
Examples of rounding(Practice)

Specified Observed To Be Rounded Conforms

Limit Value Rounded Value
Yield point,
36,000 psi, 35,940 100 psi 35900 No
min
35,950 100 psi 36000 Yes

Nickel, 57%, 56.4 1% 56 No

mass, min
56.5 1% 56 No

56.6 1% 57 Yes
Special Case-Rounding to Nearest 50, 5,
0.5, 0.05, etc.
1. Double the Value: Multiply the observed or calculated value by
2. Standard Rounding: Round this doubled value to the nearest standard
increment (100, 10, 1.0, 0.10, etc.).
3. Divide by 2: Finally, divide the rounded value by 2 to get the rounded
result.

Example: Rounding 6025 to the Nearest 50

1. 6025×2=12050
2. Round to the nearest 100: 12050 becomes 12000
3. 12000/2=6000
4. Result: 6025 rounded to the nearest 50 is 6000.
Special Case-Rounding to nearest 0.02,
0.25, 0.3, etc
Divide by the Interval: Divide the value by the desired rounding increment or
interval.
Standard Rounding: Round the result to the nearest whole number.
Multiply by the Interval: Multiply the rounded number by the desired interval.

Example 1: Rounding 0.07 to the Nearest 0.02

0.07/0.02=3.5
Round to the nearest whole number: 3.5 becomes 4
4×0.02=0.08
Result: 0.07 rounded to the nearest 0.02 is 0.08.
Guidelines for Significant Figures in Test
Results
 Ensures test results are accurate and reliable.
 Covers rules for recording, calculating, and reporting data.
 For certain purposes, such as where calculations involve differences of
measurements close in magnitude, and for some statistical calculations, such as
paired t-tests, autocorrelations, and nonparametric tests, reporting data to a
greater number of significant digits may be advisable.
 Example: Measure as 23.456 (5 digits show precision).
Recording Observed Values & Calculation
of test results
 Direct Measurements: Record all certain digits plus one estimated digit.
Example: Reading between 9.7 and 9.8, estimate as 9.76.
 Digital Instruments: Record as many digits as the device shows.
Example: Digital caliper reads 12.34 mm (4 digits).
 Don't Round Too Early: Use exact values in calculations.
 Round Final Result: Only round the final answer.
Reporting the Test Results
 Write down the standard deviation.
 Look at the first significant digit in the standard deviation.
 If that digit is 2 or higher, round the test result to that place.
 If the first significant digit is 1, round the test result to the next place.
Example:
• Test result: 3.14159
• Standard deviation (σ): 0.01
1. Identify First Significant Digit: The first significant digit of σ is 1.
2. Look at Test Result: The digit after the decimal point in the test result is 1.
3. Compare Digits: Since the first significant digit of σ is 1 and the digit after the
decimal point in the test result is also 1, we round the test result to the next place.
4. Round Test Result: Thus, 3.14159 rounds to 3.142.
Reporting the Test Results(Example 2)
 Example with σ = 0.00456:

 First Significant Digit: The first significant digit is 4.

 Rounding Rule: Since the first digit is 4 (which is greater than 2), we round to that
place.
 Rounded Result: If your test result is 2.4567, you round it to 2.457 (keeping
three decimal places).Test result: 3.14159

Example with σ = 0.0567:

First Significant Digit: The first significant digit is 5.

Rounding Rule: Since the first digit is 5 (which is greater than 2), we round to that place.
Rounded Result: If your test result is 2.4567, you round it to 2.457 (keeping three
decimal places).
Other Rules for rounding the test data

 When adding or subtracting numbers, we want the result to be as precise as the

least precise number we're working with.
• 11.24+9.3+6.32
1. Identify the Least Precise Number: Among 11.24, 9.3, and 6.32, 9.3 is the
least precise since its last significant digit is the first following the decimal
place.
2. Perform the Calculation: The exact sum is 11.24+9.3+6.32=26.86
3. Apply the Rule: We round the exact sum to the place of digits where the last
significant digit of the least precise number lies. In this case, it's the tenths
place.
4. Round the Result: So, 26.86 rounds to 26.9
Other Rules for rounding the test data
 When multiplying or dividing numbers, we want the result to have the same
level of precision as the least precise number involved in the calculation.

Example : 11.38×4.3

 Identify the Number with the Smaller Number of Significant Digits:

Among 11.38 and 4.3 has fewer significant digits (two) compared to 11.38 (four).
 Perform the Calculation: The exact product is 11.38×4.3=48.974
 Apply the Rule: We round the result to have the same number of significant
digits as the number with the smaller count. Here, 4.3 has two significant digits.
 Round the Result: So, 48.974 rounds to 49.
Reporting SD & Average
 Key Points:
1. Standard Deviation (SD): Round to two significant digits.
2. Average (Mean): Round to the same decimal place as the rounded standard
deviation.
3. Large Number of Observations: When you have many measurements, consider an
additional digit of precision.

 Steps to Follow:
1. Round the Standard Deviation:
2. Use two significant digits.
3. Round the Average:
4. Round the average to the same last place (decimal place) as the rounded standard
deviation.

 Large Number of Observations:

1. If you have a large number of observations, you might add one more digit of precision.
Examples for moderate number of
observations:
 Standard Deviation (SD): 0.05678
 Average (Mean): 2.45678
• Step 1: Round the SD to two significant digits:
SD: 0.05678 → 0.057
• Step 2: Round the Average to the same decimal place as the SD:
SD is rounded to three decimal places (0.057), so round the Mean to three
decimal places.
 Mean: 2.45678 → 2.457
 If you have more than 100 observations, consider adding an extra
digit of precision.
THANK YOU!!!!!