Technical Notes


The Mullen Burst test method (ASTM D-3786) has been with us for over half a century. It was originally developed as a quality control test procedure for the textile industry to replicate an elbow going through the sleeve of a garment. The procedure uses a smooth rubber diaphragm that is pressurized (blown up like a balloon) until it ruptures the geotextile clamped over it. The resulting value is reported in pounds per square inch (psi). While this dynamic has little correlation to the sharp angular point loads that are prevalent in geotextile applications, the mullen burst test method has lingered as a geotextile specifying tool. Currently, ASTM D-35 Committee on Geotextiles has a full portfolio of relevant test methods for a specifier to use for his specific conditions. It is important to note that mullen burst has never been adopted as a geotextile test method by D-35. An argument has been made that voids occurring between stone base material will cause the underlying geotextile to be stressed similar to the mullen burst dynamic. However, this argument can only be valid where a uniform base aggregate is used. A sound specification for road base aggregate calls for well graded angular base material where such voids do not exist.

If one believes that stresses on geotextiles used in the various applications is similar to a mullen burst test, the current AASHTO values are not substantiated by design. Both Dr. Koerner in Designing With Geosynthetics and Greg Richardson/Dr. Koerner as presented in the IFAI Design Primer have design charts for relevant mullen burst values. As Richardson/Koerner note:(1)

"Minimum burst strengths for a range of particle sizes can be estimated as shown on Figure 3.3. [see Attachment A] For aggregate size less than 2-inch and typical contact pressure, a Mullen Burst strength of only 50 lb/in2 is required to provide subgrade restraint. Most geotextiles have a burst strength in excess of 50 lb/in2 and will therefore be adequate if properly installed."
This figure is derived from the following equation:(2)
Treqd = p dv
Treqd = the required fabric strength p = the stress at the fabric's surface p, the tire inflation pressure dv = the maximum void diameter .4 da da = the average stone diameter
Looking at the test value for comparison:
Tult = Ptest dtest
Tult = the ultimate fabric strength Ptest = the burst test pressure dtest = the burst test diameter (1.2 in)
assuming that the allowable strength equals the ultimate strength,

Factors of Safety = Tallow / Treqd

FS = Ptest dtest / p dv

FS = Ptest (1.2) / p (.33 da)

FS = 3.6(Ptest / da p)

assumptions for the calculation are:

  • poorly graded stone base course (rare to nonexistent for public roads with aggregate specifications)
  • tire inflation pressure is not reduced through the aggregate layer

These assumptions add to one's factor of safety for design. This brings us back to Richardson/Koerner1 comment that "only 50 lb/in2 is required to provide subgrade restraint."

The short term and long term stresses that Richardson/Koerner are representing with their 50 lb/in2 requirement deals with the worst case situation of road stresses. At no time will such stresses be felt in drainage and erosion control applications where 75-100 psi tire loads will not be seen. Using the graph presented and assuming an additional factor of safety of 2 along with the previous factor of safety assumptions, the following mullen burst values are presented as requirements for drainage, erosion control and separation applications only where uniform aggregate is present:

Protected Drainage Unprotected Drainage Protected Erosion Control Unprotected Erosion Control Medium Survivability Separation High Survivability Separation
90 180 100 200 100 200

  • Protected Drainage assumes 2Protected Drainage assumes 2" diameter uniform stone; unprotected assumes 2"
  • Protected Erosion Control assumes a sand layer (protection layer and pore water dissipator) where 2+Protected Erosion Control assumes a sand layer (protection layer and pore water dissipator) where 2+"
  • Unprotected Erosion Control assumes more severe initial installation environment (no pore water dissipator layer)
  • Medium Survivability Separation assumes 2Medium Survivability Separation assumes 2"
  • High Survivability Separation assumes more severe installation environment

Using the available design theory shows these values to be extremely conservative for the applications noted.

It is important to note that existing mullen burst values as presented in the AASHTO M288-90 specification for drainage and erosion control (no mullen burst values specified for separation) are arbitrary. A report presented to the Transportation Research Board in 1983 showed a graph where various geotextiles' grab strengths were compared to their burst strengths. The graph produced a correlation that showed the average mullen burst value was 1.6 times the average grab tensile for these various geotextiles (both nonwoven and woven).(3) This 1.6 factor was used to determine existing Task Force 25 and AASHTO M288 values. As the paper notes,(3)

"burst strength is redundant because it is indexed by fabric tensile strength values."

The existing AASHTO M288 mullen burst values are not only redundant, but based on average product values where half did not meet the 1.6 correlation -- not a design by function approach. Thrace-LINQ, Inc. suggests a specifier not use the mullen burst test method at all; however, if you believe there is a correlation between the mullen burst dynamic and actual geotextile stresses, use the available design by function approach outlined above to determine your specification values.

(1) Richardson, Gregory, Koerner, Robert M., "Subgrade Stabilization," A Design Primer: Geotextiles and Related Materials, Industrial Fabrics Association International, First Edition, St. Paul, MN, 1992.
(2) Koerner, Robert M., Designing With Geosynthetics, Prentice-Hall, Englewood Cliffs, NJ, 1990.
(3) Weimar, Richard D., "The Mechanism of Geotextile Performance in Soil/Fabric Systems for Drainage and Erosion Control, Transportation Research Board, January 1983.