Asphalt Institute Manual Series No. 2 (MS-2) 

1. Introduction

The objective of bituminous mix design is to determine the combination of bituminous binder and aggregates that will give durability to the pavement structure. Correct mix design involves adhering to certain laboratory test procedures and design criteria and hence these laboratory methods are to be followed exactly. Any bituminous mix design procedure will have three basic components.

(a) Laboratory compaction of trial mixes.

(b) Stability and volumetric testing.

(c) Analysis of the test results.

The Marshall procedure of bituminous mix design explained here has been standardized by the American Society for Testing and Materials, given by ASTM Designation D 1559.  Several trial mixes have to be made before the mix required for the project can be finalized. To start with, a blended aggregate gradation close to the median of the specification limits is chosen for the initial trial mix. Before commencing preparation of trial mixes, it should be verified that all the mix constituents such as aggregate, filler, bitumen etc., comply to the required specifications. After carrying out the Marshall Stability and Flow tests and volumetric analysis of the trial mixes, necessary adjustments are made on the trial mix to finalise the job mix formula.

2. Aggregate

sizes and proportioning Aggregate fractions in the bituminous mix are generally designated as below.  Coarse aggregate - Retaining on 2.36 mm sieve.  Fine aggregate - Passing 2.36 sieve.  Mineral filler  - Passing 75 µm sieve. After the aggregate properties are determined and compliance with the specifications is established, one or more trial blending is prepared. The aggregate percentages are normally specified excluding binder content ie., total aggregate equals 100%. When aggregate and bitumen are mixed together, the bitumen content is expressed as a percentage of total mix. In some cases, it is also expressed as a percentage of dry aggregate.

3. Quality control testing

system for bituminous mixes Mix design is only the starting point in the process of producing a durable pavement. Field verification testing must be carried out continuously during construction, to ensure that the criteria established in the laboratory are achieved in the field also. Normally, the project quality control system for bitumen mixes will have the following 4 phases.

(a) Pre-production sampling and testing: This ensures that the constituent materials of bituminous mix viz., aggregates, filler and bitumen satisfy the physical requirements set forth in project specifications.

(b) Job mix formula verification: Tests are done at the start of the plant production to compare the field mix with the approved job mix formula. Some slight adjustments may be required due to the variations in aggregate characteristics. This adjusted job mix formula may remain as the target for all subsequent quality control testing.  

(c) Daily job control testing: Representative samples of hot mix are taken regularly and subjected to quality tests. The results are then compared with job mix specifications or requirements.

 (d) In place acceptance testing: The sampling and testing of in-place bituminous layers are carried out in the last phase. The acceptance of paving work is based on the result of this.

4. Volumetric properties of compacted mix

In order to determine the volumetric properties of the bituminous mix, specific gravities of the constituent materials are to be found. The specific gravity of aggregates is computed in three ways:

(a) Bulk specific gravity.

(b) Apparent specific gravity. (c) Effective specific gravity.

4.1 Bulk specific gravity of aggregate (Gsb)

Suppose P1 and P2 represent coarse and fine aggregate having bulk specific gravities G1 and G2 with 60 % and 40 % percentage compositions by weight in a mix. Then for 5% binder content by total weight of the mix, the total aggregates will be 95% in the mix. The percentage weight of aggregates will be 57 and 38 (0.95 times individual percentages) in the total mix. Now the bulk specific gravity of the aggregate mix is computed as  Gsb = 95/[(57/G1) + (38/G2)] In determining the bulk specific gravities of aggregates, the weight of saturated surface dry (SSD) aggregate is used for computing the volume.

4.2 Apparent specific gravity of aggregate (Gsa)

Apparent specific gravity (Gsa) of an aggregate mix is calculated in the same way as for effective specific gravity (Gse) except that apparent specific gravities of individual aggregates are substituted in the formula. In determining the apparent specific gravities of individual aggregates, oven dry weight of samples are taken to compute the volume - assuming that the material is impermeable.

4.3 Effective specific gravity of aggregate (Gse)

This includes all voids except those that absorb bitumen. Effective specific gravity (Gse) is calculated as below.  Gse = (Pm - Pb) /[(Pm/Gt) – (Pb/Gb)] where  Pm  = Percentage weight of total mix = 100 .  Pb  = Percentage of binder by total weight of mix.  Gt  = Theoretical maximum specific gravity of mix (no voids).  Gb  = Specific gravity of bitumen.

4.4 Maximum specific gravity of the compacted mix (Gt)

The maximum specific gravity or the theoretical specific gravity (Gt) is determined by ASTM D 2041 or AASHTO T 129 method. Gt is the ratio of the weight of the bituminous mix in air without any voids to the weight of an equal volume of water. A vaccum procedure is used to remove the entrapped air in the mix and determine the volume in a void less state. For each measured Gt, the Gse can be computed using the formula given in 4.3. Using the average of Gse values, the Gt for any binder content can be computed as below.

 Gt  =Pm /[(Ps/Gse) + (Pb/Gb)] where

Pm  = Percentage weight of total mix = 100.

 Ps  = Percentage of aggregate in the mix (100 minus % of binder by weight of total mix). 

Pb  = Percentage of binder by total weight of mix.  

Gse  = Average effective specific gravity of aggregate.

Gb  = Specific gravity of bitumen.

Example:  If Gt of a bituminous mix is 2.535 as per ASTM D 2041 determination and Gb = 1.030 then Gse for 5.3 % binder content will be 2.761.

With Gse = 2.761, Gt for 4 % binder can be computed as below.  Gt  = 100 /[(96.0/2.761) + (4.0/1.030)] = 2.587.

4.5  Bitumen absorption

Mineral aggregate is porous and can absorb bitumen as well as water. The percentage of bitumen absorbed (Pba) by the aggregates in the mix is expressed as a percentage by weight of aggregate (not as a percentage of total mix). Pba is calculated as below.

 Pba  = 100 x [(Gse – Gsb)/(Gsb x Gse)] x Gb where

Gse  = Effective specific gravity of aggregate mix.

Gsb  = Bulk specific gravity of aggregate mix.

Gb    = Specific gravity of bitumen.

4.6 Effective bitumen content for a bituminous mix (Pbe)

Pbe is the total bitumen content less quantity lost by absorption into aggregate. It is the bitumen content that remains outside the aggregate particles as a coating. It is Pbe which governs the performance of a paving mix.

Pbe    = Pb – [(Pba/100) x Ps] where

Pbe   = Effective bitumen t content, % by weight of total mix.

 Pb  = Percentage of binder by total weight of mix.

Pba  = Percentage of absorbed bitumen, % by weight of aggregate

 Ps  = Percentage of aggregate in the mix (100 minus % of binder by weight of total mix)

 Example: If binder content is 5.3 % and assuming absorption of 0.8 %, then effective bitumen content is  Pbe  = 5.3 – (0.8/100) x 94.7 = 4.5 %

 4.7 Voids in mineral aggregate for a compacted mix (VMA)

VMA is expressed as a percentage of the bulk volume of the compacted mix  

VMA = 100 – (Gm /Gsb) x Ps where Gm = Bulk specific gravity of compacted mix (computed from     saturated surface dry weight of compacted specimens).

Gsb  = Bulk specific gravity of aggregate mix (Using formula in 4.1)

 Ps = Percentage of aggregate in the mix (100 minus % of binder by weight of total mix)

 4.8  Percentage air voids in compacted (Va)

Va = 100 x (Gt – Gm)/Gt where  

Va = % air voids in the mix by percentage of total mix.  

Gm = Bulk specific gravity of compacted mix.

 Gt = Maximum specific gravity of mix (By direct measurement by ASTM D 2041 method or calculated using formula in7.1.4).

4.9 Percentage voids filled with bitumen in the compacted mix (VFB)

VFB is the percentage of VMA filled with bitumen and is given by  VFB = 100 x (VMA - Va)/VMA where

VMA = Percentage of air voids in aggregate of bulk volume. 

Va  = % air voids in the mix by percentage of total mix by volume. A typical work sheet for computing the density void analysis is given as form QC-B5/2014.

5. Preparing bitumen mix in Marshall Method of mix design

5.1 General The original Marshall method is applicable only to hot bituminous paving mixes, with a maximum aggregate size of 25 mm. Subsequently modified Marshall Method has been developed for aggregates with maximum size up to 38 mm. The Marshall method uses standard test specimens of 64 mm height and 102 mm diameter. They are prepared using a specified procedure for proportioning materials, heating, mixing and compacting the aggregatebitumen mixture. The two principal features of this method are a density void analysis and a stability flow test. The stability is the maximum load taken by a specimen when tested after keeping in 60°C water bath for 30 to 40 minutes. The flow value is the deformation of the specimen in units of 0.25 mm or 1.0 mm at the maximum load condition. Marshall Stability test is empirical in nature. Hence no modifications can be effected to the standard procedure, likely reheating of mix for preparing specimens.

5.2 Initial binder content for test specimens

To provide adequate data, at least three specimens are prepared for each bitumen content. The bitumen contents for trial mixes are selected in increments of 0.5%. At least two shall be below the expected design value or specified value and two above this. So, if six binder contents are tried, then 18 specimens will be required. Each specimen requires about 1200 gm of aggregate. The minimum aggregate requirement for 1 set of test will be 23 kg with 4 litres of bitumen. If no binder content is specified, the expected design binder content can be arrived using the following formula.

P = 0.035 a + 0.045 b + K c + F where

 P = Approximate % binder content by weight of total mix

 a = Percentage of aggregate retaining in 2.36 mm siev

 b = Percentage of aggregate passing 2.36 mm sieve and retaining in 75 µm siev

 c = Percentage of aggregate passing 75 µm sieve K = A factor depending on percentage of aggregate passing 75 µm (No. 200) sieve    = 0.20 % for 5 % passing 75 µm sieve    = 0.18 % for 6 to 10 % passing 75 µm sieve    = 0.15 for 11 to 15 % passing 75 µm sieve F = A factor based on absorption of aggregate, varying from 0 to 2.0. Normally it is taken as 0.70. 5.3 Preparation of test specimens The following steps are recommended for preparing Marshall Test specimens: At least three specimens are required for each aggregate grading and bitumen content. Coarse and fine aggregates to be incorporated into the bitumen mix are initially subjected to various tests to ascertain its suitability. The following tests are generally essential.

(a) Specific gravity/water absorption.

(b) Los Angeles abrasion/Impact value.

(c) Sieve analysis.

(d) Sulphate soundness. 

(e) Flakiness and Elongation.

(f) Stripping. The selected aggregates are dried to constant weight at 105oC to 110oC. The aggregates are then dry sieved. The proportioning can be done by trial and error method. An initial gradation very close to the median of specification limits is selected for preparing specimen

5.4 Selection and testing of binde

 The selection of the bitumen grade for any paving mix is done based on the temperature, rainfall and traffic of the construction zone. The selected bitumen is subjected to the following tests

 (a) Penetration.

(b) Viscosity

 (c) Specific gravity.

(d) Ductility.

(e) Solubility.

(f) Softening point.

5.5 Mixing and compaction temperatures

The temperatures up to which the bitumen is to be heated to produce viscosities of 170±20 centistokes kinematic and 280±30 centistokes kinematic are denoted as mixing and compaction temperatures for the paving mix. By plotting viscosities against temperatures in a log scale, these values can be obtained.

5.6 Preparation of mould and hammer

The mould assembly is cleaned and heated to a temperature between 95oC and 150oC. A piece of filter paper, cut to size is placed at the bottom before the hot mix is poured.

5.7 Preparing the bituminous mix

Weigh into separate pans for each specimen, the required quantity of different aggregates. This will be about 1200 gm. It is desirable to prepare a trial specimen to work out the weight of material required to produce a specimen of height 63.5 mm. If the height of the trial specimen falls outside the limits, the quantity of aggregates has to be adjusted as below. Adjusted weight of aggregate = 63.5 mm x   Weight of aggregate used              Height of trial specimen in mm  Heat the pans containing the aggregate mix in the oven to a temperature not exceeding 28oC above the mixing temperature determined as per 5.5. Transfer the aggregate mix to the mixing bowl and agitate thoroughly. Add the required weight of heated bitumen to the aggregate mix. Preferably with a mechanical mixer or by hand with a trowel, mix the aggregate and bitumen together till a uniform and well coated mix is obtained.  Place the mould assembly in the compaction pedestal. Place a paper disc at the bottom of the mould. Transfer the mix into it. Spade the mix with a heated spatula or trowel 15 times around the perimeter and 10 times over the interior. See that the temperature of the mix is not below the compaction temperature as per 5.5

5.8 Compaction of specimens

Place a paper disk on top of mix. Apply the number of blows depending on the design traffic category, using the compaction hammer with free fall of 857 mm.

Apply the same number of blows after reversing the specimen. Keep the mould with specimens to cool, normally for overnight. Remove the specimens using extrusion jack and keep on a smooth level surface ready to test. 

6. Test procedure for Marshall Method of mix desig

6.1 General

 In the Marshall method of mix design, each compacted test specimen is subjected to the following tests and analysis. (a) Bulk specific gravity (Gm) determination. (b) Stability and Flow test. (c) Density and void analysis.

6.2 Bulk specific gravity (Gm)determination

Bulk specific gravities of saturated surface dry specimens are determined as per ASTM D 2726. This test is performed according to ASTM D 1188 for paraffin-coated specimens.

6.3 Stability and flow tests

After determining the bulk specific gravity of the test specimens, the stability and flow tests are performed. Immerse specimen in water bath kept at 60oC±1oC for 30 to 40 minutes before testing. When the testing apparatus is ready, remove the specimen from water bath and carefully dry the surface. Place the specimen centrally on the lower testing head and fit the upper head carefully. Fix the flow meter with zero as initial reading. The load is applied at a constant rate of deformation of 51 mm per minute. The total load at failure is recorded as its Marshall Stability value. The reading of flow meter in units of 0.25 mm gives the Marshall Flow value of the specimen. Flow value is reported in mm also. The entire testing process, starting with the removal of specimen from bath up to measurement of flow and stability, shall not take more than 30 seconds. While the stability test is in progress, hold the flow meter firmly over the guide rod, remove when load begins to decrease and take the reading and record.

6.4 Density and voids analysis

After completion of the stability and flow test, a density and voids analysis is done for each set of specimens. The method of calculations is given in article 4 above – volumetric properties of compacted paving mix. (a) Average the bulk density determinations, for each bitumen content. Values obviously in error need not be considered. This average value of Gm is used for further computations in void analysis.

(b) Determine the theoretical or maximum specific gravity (Gt) by ASTM D 2041 method for at least 2 bitumen contents nearer to the optimum binder content. An average value of effective specific gravity (Gse) of aggregate mix is found using these values. This average Gse can be used for calculating Gt for different bitumen contents.

c) VMA, Va, and VFB are then computed using the equations given under articles 4.7 to 4.9 above.

7. Analysis of Marshall Mix design data  

7.1 Preparation of test data The stability/flow values and the void data are prepared as below.

(a) The observed stability values are corrected multiplying by a factor, if the thickness of specimens is not 63.5 mm. Table 1 gives the correction factors.

(b) Average the flow and the corrected stability values. The values obviously in error can be discarded.  (c) Prepare the following graphs with bitumen content in X-axis:

(i) Stability Vs bitumen content. 

(ii) Flow Vs bitumen content.

(iii) Unit weight of total mix (Gm x 1000) Vs bitumen content.

(iv) Percentage air voids (Va) Vs bitumen content.

(v) Percentage voids filled with bitumen (VFB) Vs bitumen content.

(vi) Percentage voids in mineral aggregate (VMA) Vs bitumen content. These graphs are used to get the design bitumen content of the mix. 

Table 1: Correlation values for correcting observed Marshall Stability values 

Nominal maximum


Minimum VMA %


particle size


Design air voids %














































NOTES: (i) As per MS-2, the nominal max particle size is one size larger than the first sieve to retain more than 10 percent.

(ii) Interpolate for VMA for air voids between the given values.

7.3 Trends in mix design

 data  The curves plotted with test data prepared as mentioned in

7.1 (c) are attached for guidance

 Normally the curves follow a general pattern even though variations can occur. The trends generally noted are:

(a) The stability value increases with increase in bitumen content up to a maximum point and thereafter it decreases.

(b) The flow value consistently increases with increase in bitumen content.

(c) The unit weight (Gm x 1000) of the mix increases with increase in bitumen content up to certain limit and then it decreases.

(d) The percentage air voids (Va) steadily decreases with increase in bitumen content. (e) The percentage voids in mineral aggregate (VMA) decreases initially and then it rises

 (f) The percentage voids filled with bitumen (VFA) steadily increases with increase in bitumen content. This is becau

e the VMA is being filled with bitumen

8. Selection of final mix design

8.1 General

Mix should not be designed to optimize one criteria. Mixes with abnormally high values of stability are found to be less durable and such mixes tend to crack prematurely. Normally the range of bitumen content, which passes all the criteria, is very narrow. Bitumen content within this range has to be chosen by the designer for the specific project. Some of the considerations for selecting the final bitumen content are discussed below.

8.2 Voids in mineral aggregate (VMA)

The VMA in the aggregate mix is not constant. With the addition of bitumen, the mix becomes more workable and compacts easily, which shows an increase in bulk density and decrease in VMA in the initial stage. The VMA curve has got a flattened U-shape as seen from the data curves. At some point when bitumen content increases, the VMA increases. Any binder content beyond the minimum point must be avoided. A mix with bitumen content on the “wet” side or right hand side of this curve has a tendency to bleed when placed in the field. When the bottom of the U-shaped curve falls below the minimum VMA criteria, for the maximum size of aggregate used in the mix (Table 3), it indicates that a change in job mix formula is essential. The aggregate grading is to be changed to provide additional VMA. If the entire curve falls below the minimum line, a significant revision of the mix including changing the source of materials will be required.

8.3 Effect of air voids (Va)

The design air voids (Va) of 3% to 5%, is the level desired after several years of traffic. This does not vary with the traffic volume in these years. The compactive effort has to be selected according to the volume of traffic expected. A pavement mix that ultimately consolidates to less than 3% air voids is expected to rut and shove. In the other way, if the pavement is constructed initially with more voids, brittleness, premature cracking, raveling, and stripping can happen. If the right compactive effort is chosen and if the percentage air voids after construction is about 8%, the design air voids range will normally be achieved.

8.4 Effect of voids filled with bitumen (VFB)

The criteria, Va, VMA and VFB are interrelated. If any one of the two values is within the limits, normally the third will be also in order. The VFB criteria help to limit the maximum bitumen content, so that a minimum air voids is included in the mix composition.

8.5 Effect of compaction

level For the same bitumen content, both air voids (Va) and voids in mineral aggregate (VMA) decrease with higher compactive effort. The bitumen content value for minimum VMA decreases with increase in compactive effort. So if a mix is designed with 50 blows (medium traffic) and if the pavement has to bear heavy traffic, the selected bitumen content will be on the higher (wet) side of VMA curve for 75 blows compaction. Ultimately a mix susceptible to rutting will be the result. This can happen in the other way also. A mix designed for heavy traffic used on a pavement to carry out light traffic will end up with a high percentage of air voids (Va). Such a pavement will crack prematurely or even the aggregate will ravels out due to loss of bitumen adhesion. So it is important to note that the compactive effort selected in the laboratory should simulate the expected design traffic.

 8.6 Influence of structure and climate

Mix design is a compromise of several factors. A mix that provides best overall performance when used in the pavement should be the one with the design bitumen content. For a bituminous concrete overlay over a cement concrete pavement, the main consideration will be to limit the rutting and minimize the reflective cracking. Mixes with bitumen content on the higher side of acceptable range are avoided in such situations. However, if the sub grade is not adequate the pavement performance has little to do with the mix design. With all other factors being equal, mixes with bitumen content on the higher side of the range are less prone to cracking. Similarly, mixes on the low range of bitumen content are less susceptible to rutting. In some cases, the bitumen content may be selected depending on the future maintenance concern. Finally, climate can also have an impact on the performance of the mix or the pavement. Mix design does not consider this factor except in selecting the grade of the bitumen. In hot climates, harder, more viscous binders are normally used to obtain more stability. Bitumen content on the lower side of acceptable range is recommended in these areas. In colder climates, softer, less viscous binders are recommended to produce a mix, which is less susceptible to low temperature shrinkage cracking. Rutting is less of a concern and so bitumen contents on the higher side of acceptable range are recommended to furnish a mix, which is more elastic and resilient. However, it shall be kept in mind that the shift in bitumen content is only a minor amount within the narrow range that passes all the mix design criteria. Traffic should be held off the pavement as long as possible while the mix is cooling to normal temperatures. This will help to impart more stability to the mix laid.