Farhad Salour Doctoral Thesis

8.1.
 
Assumptions and limitations 
The environmental data collected during the study period were limited to certain 
locations throughout the test section. However, the temperature and moisture profile 
measurements were assumed to be identical throughout the whole test section. A 
similar assumption was made for the deflection measurements. Since the pavement 
structure at the test road section was in-service, a number of the FWD measurements 
were limited due to traffic safety regulations. This study was to a considerable extent 
based on data collected by in situ measuring instruments and their accuracy level. The 

---

53 
findings from this study are limited to findings from a certain pavement structure and 
test site and therefore cannot be generalized. It would be beneficial to compare the test 
results from this study to results from other pavement structures and test sites. For the 
laboratory-based investigations, it is believed that incorporating soil suction into 
predictive resilient modulus models appears to give a new insight into predicting 
seasonal resilient modulus variation due to moisture content changes. However, further 
validation across a wider range of materials may be required, for both the approach and 
the recommended parameters. 
8.2.
 
Recommendations for future work 
From the research carried out in this thesis, there are still many questions that need to 
be further investigated in future research studies. The field study carried out here was 
limited to a single specific test site with specific properties. The response and the stress 
dependency of the unbound layers and their correlation to the measured moisture 
content were evaluated using backcalculation of the surface deflection data. Direct 
measurement of in situ stresses and strains using pressure and deformation gauges along 
with moisture sensors (or even suction measurement sensors) under moving traffic 
loads and their response to seasonal moisture variation would be an interesting 
question, which should be further investigated. 
Considering the importance of permanent deformation in unbound pavement layers in 
rut formation, particularly in thin flexible pavements, it is evident that more studies 
need to be carried out on the permanent deformation properties of unbound materials 
and the influence of the moisture content. Even though only two subgrade soils were 
tested, it was observed that in most of the current permanent deformation models (due 
to multiplicity of the model parameters) usually several combinations of parameters 
would provide acceptable fit to the test data. This can cause a major confusion for a 
user to select the right set of parameters when trying to predict the permanent 
deformation of the material in question. This major drawback is probably due to lack of 
sufficient permanent deformation studies There is an apparent need and is 
recommended that more permanent deformation tests be conducted to set up widely 
accepted range limits for the parameters of different models and if possible develop 
simpler models with a minimum number of parameters. 
Moreover, more research needs to be carried out to calibrate and implement permanent 
deformation models that are based on data from multistage RLT tests. Comparing the 
laboratory-based model predictions against the in situ pavement data or controlled APT 
tests would be recommended as part of the evaluation and validation of the models. 
Finally, and as the field and laboratory measurements illustrated, the environmental 
factors and their effect on pavement material behaviour and performance are very 
complex in nature and it is believed that more research studies are certainly required to 
progressively obtain a more complete understanding of the phenomena involved 

---

54 
through both laboratory and field measurements. This will assist in developing 
mechanistic approaches that can quantitatively account for climatic factor effects on 
material behaviour and pavement performance. 

---

55