Heavy Duty Pavement Design (Flexible) Over Soft Subgrade 1

[MAKP]

I am looking for advice on design methods for heavy duty pavement over weak, compressible subgrade conditions. Due to the likelihood of ongoing differential settlement at the site, I am considering a flexible pavement design (asphalt over granular base). The subgrade consists of sand fill overlying peat and organic silt.

MAKP

 

[dik]

If subject to heavy loading (heavy duty?), then an asphaltic pavement would more likely 'break up' under loading. Using a geotech membrane beneath the granular base may mitigate it somewhat, depending on how poor the sub-base is. Life cycle on it would likely be short.

 

[VAD]

Hello MAKP:

Heavy duty pavements have been built over soft subgrades and have performed satisfactorily for the conditions you have described. Settlement/differential settlement would perhaps be the most worrysome problem. Preloading, if feasible, and delayed construction of the pavement structure have been used with good success. The depth of soft ground would be important to know in this regard.

While a rigid pavement structure is assumed to bridge softer ground I would be concerned about whether settlement is an issue at this time since this would also influence the behaviour of the rigid pavement. I presume that your choice of a flexible pavement has therefore to do with deepseated settlement. Generally, it is often easier and cheaper to remedy a settling asphalt pavement than a broken up concrete pavement to provide satisfactory rideability

Geotextiles, geogrids etc do not stop vertical settlement in deep settling deposits but are somewhat advantageous in minimizing lateral spread.

What is the experience in the area. You may wish to research this as well.

 

[MAKP]

Thanks for the responses; however, I am looking for advice on the design approach to the problem. For example, is the most appropriate design method a linear elastic layer model such as KENLAYER. How about the British Port's method? I had heard that the ASCE had a committee that was going to develop a manual for pavement design for ports but I have never been able to find out whether or not it has ever been published.

Answering some of the other questions - the site has a varied loading history and has undergone considerable preloading in some locations and little if any at other locations (it is a very large site). To futher complicate matters, the thickness of existing fill over the site varies from less than 1 m to more than 5 m. The peat thickness varies from < 0.3 m to as much as 3 m (variation is both natural and due to compression/consolidation). Underlying the peat is organic clayey silt to clayey silt. The highly compressible peat and silts extend to depths ranging from 10 to 15 m. As expected, the moisture content in the peat is lower in locations where the overlying fill is thicker.

 

[VAD]

Hello MAKP:

The desigh approach to your problem does not lie with any of the pavement design models you have mentioned since most if not all design methods assume that the underlying ground below the pavement subgrade is &quot;well behaved&quot; in the sense that there are no serious or complicating geotechnical issues. As such the KENLAYER or Ports method etc would not be of assistance except for the determination of the layers etc above the subgrade.

The problem that you face is best resolved by understanding your foundation conditions in relation to settlement and differential settlement.

You have to look at fundamental aspects of foundation behaviour which will occur when you load the site with a pavement structure. Look at stress distribution that will be invoked by pavement and load conditions. What will be the likely depth of influence of the superimposed stress due to traffic loading etc?. Does it still matter if I have 10m of soft ground. Is my ground still consolidating? What is the likely impact of the increased stress of further ground consolidation and what is the zone of influence and so on etc.

Fundamental ground investigation information needs to be mulled over to determine based on experience and judgement what is likely to happen. Is it posssible to still preload the areas that were not preloaded?. This would remove some of the undesirable features of differential settlement.

Is it possible to excavate. This is often shied away in peat areas but has been used with success. Your silty subsoil my cause you a problem in this respect, but can be done if needed.

You will probably end up floating the pavement on this ground and could expect to have maintenance problems. Can this be lived with. This may have to be lived with as other approaches may be cost prohibitive. Such is life in dealing with these deposits and ground conditions.

The decision for a job of this sort brings together a series of knowledge and generally can only be best resolved by someone being in intimate contact(hands on) with the site and conditions, proposed development re parking areas etc etc.

This is a job for someone who is grounded with both geotechnical as well as pavement design experience. Unfortunately, in today's practice these are often relegated to separate camps and hence the uncertainity of what can happen and hence indecisions.

You have a challenging problem and I wish you the best.

 

[jdmm]

You can use the ASHTO resilient modulus approach for heavy pavements but first you have to improve the subgrade sufficiently to apply the method. For pavements over peat we normally preload to improve the strength and stiffness. Presumably if it is a peat environment the grades are relatively low and would have to be increased. This allows the placement of permanent fill. Preloading will cause settlement so that additional permanent fill will result. We have found that for normal truck loading you have to have at least 1.2 m (4 feet) of improved subgrade before the pavement so that the wheel stresses at the level of the peat are within tolerable limits.

For something like a container terminal where the loading from container lift trucks can be substantially higher, a thicker subgrade would be required. A simple approach is to look at the increase in stress at the base of the fill from the wheel loading and compare it to the strength of the underlying material. Geogrid can be used to reduce the total thickness of fill required to protect the natural subgrade from overstressing but I haven't found a method of analysis that I am comfortable with yet so our tendancy is to stick with thick mineral fills over the peat.

Depending upon the requirements of the pavement (ie storage of containers) a preload of the order of 5 feet for vehicle travel is usually sufficient. If you are storing heavy loads you should use a preload of about 1.5 times the anticipated load on the pavement. For a peat thickness of about 10 feet a 5 foot preload can cause settlement of the order of 2 to 3 feet which will become settlement fill. You have to add the anticipated settlement fill to the preload fill so that the required preload can be removed. This would mean that for the 5 foot preload and 2.5 feet of settlement you would have to put on 7.5 feet of fill which would cause more settlement. You have to iterate through the calculations to find the ultimate fill height required to remove 5 feet of preload. You need to have good information on the distribution of compressible settlements to do this properly because thicker fills will be required over more compressible sections and thiner under less. Otherwise you end up taking off too much fill in some areas and not enough in others resulting in a non-uniform preload of the underlying soils.

Secondary consolidation of high moisture content soils and peat is also a concern since preloading will not eliminate post construction, secondary consolidation, it will only reduce it.

 

[AZdinak]

I tend to agree with VAD. Pavements can be designed for poor subgrade support characteristics (CBR values less than 1) by using adequate thicknesses of stone subbase and asphaltic base in conjunction with subgrade reinforcement with grids (bi-axial). The more interesting aspect of the problem comes from the long term distress caused by compression of the subsoils. Pavement design limits stresses on subgrades to allowable values during anticipated &quot;fairly transient, live loading&quot; conditions (if assuming traffic). However, the pavements are just as suceptible to distress due to primary consolidation and continued secondary compression. If you must limit the secondary compression, we have applied Ladd's AAOS methodology and designed our surcharging to limit calculated secondary compression values to less than several inches in a reasonable design life. This approach, of course, assumes that the surcharge is maintained for a period of time long enough for &quot;total settlement&quot; to be less than the threshold value (including immediate, primary, and secondary).

If you are looking at container storage, you likely need to get your settlement taken care of before paving or anticipate recommendeding funds for maintenence.

Zdinak

 

[BigH]

Having quickly scanned the comments - VAD appears to have hit it on the head. The problem is not with the pavement design. It is with the foundation on which the pavement is resting. This happens many places. On the project I am currently working we have soft to firm clays and will be placing embankments on them up to about 4 to 5 metres in height. The pavement structure (asphalt, base course, subbase, subgrade) is not the critical part other than the thickness has to be designed for the loadings and number of expected axles in the time frame. But, with this, the foundation on which the embankment is built will be settling some 100 to 200mm over the years. In this case there is very little you can do unless you want to preload all the area, use wick drains, etc. so that all the settlement is built out. Settlements will happen. Just remember to try to build out the settlements at the bridge ends. Perhaps overbuild the embankment at the bridge ends, leave it on for as long as you can before you remove and build the pavement structure - this way you will at least minimize to the extent practical the settlments that happen there. Also remember that if the embankments settle evenly, the pavement structure won't really know it I opine. If you go from areas that have been preloaded earlier, then not, then, then not - well, you'll have a bit of rollercoaster but this isn't likely all that critical. I would tend to doubt that you will have large differential settlements in short lengths so problems associate with this will be less.
Now if you have such major settlement issues, you won't need to overcompact the embankments. The small settlements that would occur if compacted to 90% MDD vs 95%MDD (standard or modified) is not likely significant compared to the foundation settlement.

 

[MAKP]

As some of you have mentioned, preloading will reduce but not eliminate post-construction settlement. I always use this phrase when dealing with developments over peat and soft soils. The performance of the pavement will be highly dependent on the diffential settlement. Knowing the loading history of the site and the stratigraphy, I can predict where to expect problems due to differential settlement.

Preloading, wick drains, prediction of settlement and bearing capacity these are all geotechnical issues I deal with on a daily basis and have a considerable amount of experience. I am less experienced with heavy duty pavement design.

I have done a lot of reading about pavement design but most of the methods are based on standard 18 kip axle loads. Rutting models look at a limiting compressive strain at the top of the subgrade; however, the stress distribution with depth for a standard axle loading is considerably different than from the front axle of a reach stacker.

Again I ask - are there any design methods out there based on heavy equipment that might be applicable?

Re: Geogrid

I have used geogrids in unpaved roads such as mine haul roads and log sort yards but I am not considering them for this project because I expect the pavement would rut long before the strength of the geogrid was mobilized.

I would use a geotextile as a separator in areas where there is little or no sand fill overlying the peat.

 

[VAD]

Hello MAKP

I think the Asphalt Institute has a publication on Heavy Duty Pavement Design in one of their series. You may wish to check their web site. If I remember correctly they have examples with loaders etc.

 

[Ron]

MAKP...there have been a lot of good suggestions here that will likely help you. I've designed similar pavement structures (for paper mills in the SE USA in swamps!). One thing to consider is the premise that you will need a lot of high modulus material to attenuate the loads, but high modulus materials are typically dense and heavy, thus increasing the potential for long term settlement. Consider using a foamed concrete fill such as Elastizell. This can be designed with relatively low unit weights thus decreasing the fill load, while having a relatively high unconfined compressive strength (300-400 psi).

You can optimize the material to reduce costs and you can utilize some soil improvement techniques in conjunction to give better results.

As for analyzing your issues, I would suggest you go beyond the AASHTO methods and look at elastic layer analysis or a finite element approach.

 

[BigH]

Ron: glad to see someone else uses (has used) Elastizel. I used it way back in 1983 in the Vancouver area. It has its good points - but it is expensive so one must, in my view, pick and choose its application appropriately. You could use hogfuel or sawdust (encapsulated) for lightweight fill, then have some &quot;room&quot; for the heavier pavement structure - use geogrid to reduce a bit of the required granular subbase/base.

Before I forget - Happy Diwali to all our Indian colleagues!

 

[g7mann]

g7mann:

Hi - this use of heavy duty pavements over "soft" and/or compressible soils is difficult and challenging. You might look at Geofoam, a lightweight polystyrene material as an underlay for a road [pavement] section. This allows for overexcavation and removal of up to several feet of in-situ material, a depth determined by the ultimate traffic loading, and its' replacement with a lightweight but structurally competent material. Geofoam weighs between about 1/2 and 1 pcf, considerably lighter even than [pure] peat. "Geofam Geosynthetic" by John Horvath might be a useful text, as can be the Syracuse University [New York] geofoam web page.

Also, "Guidelines for use of Fabrics in COnstruction and Maintenance of Low-Volume Roads by Steward, Williamson & Mohney prepared by the U.S. Forest Service and the U.S. Department of Transportaion may be of assistance.

Good Luck!