Asphalt Surface Variations 1

[skillz]

A street that was just paved under my company's supervision seems to have some asphalt surface variations that make the ridability somewhat bumpy. Both the consultant(us) and the Village are questioning what the contractor could have done wrong to make this road feel so bumpy.

The contractor denies that they did anything wrong and is blaming the use of a 2.5" lift of surface course on the poor ridability.

The road was a partial reconstruction with spot curb repairs. We believe that the curb repairs may have something to do with the ridability, but the curb profile isn't showing this to be the major reason here. It seems to be more of a paving issue.

The road was constructed with 3/4" of Polymer Level Binder and then the 2.5" Superpave N50 Surface Course lift. The contractor used a 40 foot ski when paving to maximize pavement smoothness.

Does anyone have any ideas of what happened here?

We've never dealt with this problem before. Any other pavement surface problems have been identified before by the inspector in the field, but the inspector here did not see anything out of the ordinary.

 

[Ron]

How was the base prepared? The surface will reflect issues with the base, particularly surface variations.

You might also want to check the asphalt density in a "high" area and then in a "low" area to see if there is a density contribution to the issue.

Next check the paving equipment. If screeds were not tight and allowed to drift, this causes rideability issues.

Next check the laydown temperature and the time for breakdown rolling. If the temperature was variable, you'll get variations in the asphalt's response to breakdown rolling.

Did they use a traffic roller after the initial compaction and if so, how long after? If not, why not?

 

[skillz]

Ron - I don't know all of the details since this wasn't my project, but I do know that the base was existing asphalt and they did do a 3.25" grind of the existing pavement so that they could put back a new 3.25" pavement section.

One of my first thoughts was that the grinding operations may have taken out more than 3.25" in some areas and the surface would have been placed thicker than 2.5" at those areas. That may have caused this problem...

 

[cvg]

This should be the contractors problem and specs should require his construction to meet certain tolerance. Can't blame the asphalt for laying itself down wrong. Did he meet the required construction tolerance?

At this point, you should be either a)ask the contractor to repair the road, b)accept the road as is. Of course, you may want to wait 9 - 10 months till his warranty period is almost up and then inspect the roadway again at that time to decide what to do.

 

[LCruiser]

You blew it by just patching a few old curb sections. You should have replaced the curb too and run a new profile, or advised the client a smoothness spec coudn't be met. (unless the old curb met it???) Your bandaid is probably non-defensible.

 

[proletariat]

Can you be more descriptive about 'asphalt surface variations'? Also, how long ago was the reconstruction project done?

 

[BigH]

One thing that I have seen before is out of sync vibrations of the paving machine's initial packing unit. In sections of asphalt before it was noted, we ended up with a "wash board" effect. As Ron says - poorly prepared base onto which the pavement will be placed - whether it is a wash-boarded granular base, or poor patching of potholes or, in your case, kerbs, will cause a bump to reflect through such a thin pavement layer.

 

[Ron]

skillz...it's possible that what you described happened as well. The milling operation can be controlled quite well with respect to grade and milling depth; however, if the original base undulated, the milling machine will skim off the top of the base in high spots.

Another common reason for this is intermittent milling; where sections are milled and other sections are not.

 

[moogbro]

A bumpy ride can be caused by any number of issues. Typically specifications include a straight edge spec or a profilograph spec to describe what level of bumpiness is acceptable. If you don’t have this spec, you are likely stuck with a bumpy project.

Before you have a chance of figuring out what went wrong, you first must have a concept of how a paving machine works, than observe how the work is being performed, and correct it before it’s an issue. Attempts to figure it out now will likely just be conjecture.

What a lot inspectors/engineers and even some contractors don’t understand with respect to paving is how a paving machine actually works. A paving machine simply pulls a screed, the hot asphalt concrete material slides under the screed and is spread out. There are no physical forces pushing down on the screed other than it’s own weight (aka “the floating screed”).
(see

http://en.wikipedia.org/wiki/Free_floating_screed)

.

As it is being pulled, there are many forces acting on the screed which will make it go up and down and cause a bumpy mat. Obviously you don’t want a bumpy ride, so the ideal (perfect paving) situation would be to keep all the forces constant (paver speed, amount of material fed to the augers, the distribution of the material in front of the screed, temperature of material, etc). Additionally the screeds tow points need to follow along a theoretical line a constant distance above the theoretical plane you are trying to pave. Pavers with automatic controls are designed to raise and lower the screeds tow points to keep them on theoretical line as the tractor portion pulling the screed rises and sinks in holes. The 40’ ski your contractor used skimmed across the high points, and the automatic controls should have been holding the tow point(s) a constant distance above a reference point in the middle of the ski.

One additional point to be made is that assuming that the screed spreads the material across a perfect plane, what happens after the rollers hit it. As the volume before compaction is 20% to 25% more than after the rollers hit it, thicker areas will compact more than thinner. A general rule of thumb is the more lifts you can place the smoother your ride will be. If what you are paving over is rough, than under ideal paving (everything constant, with the best automatic controls) you can only expect your ride to be about about 70 to 75% better. If you can place it in two lifts, than the best you can get is ~95%. If your surface you were paving over were really rough, the best you could hope for would be to have it about 30% as rough. This general rule is why legitimate specifications limit the use of a profilograph on projects where only two lifts using automatics can be placed,not to mention require the use of a paver with automatics and a minimum ski length.

If you want a good reference, buy the “Compaction of Hot Mix Asphalt Pavements” Manual published by the Asphalt Institute.

With respect to your inspector, I can’t tell you how many I have encountered that don’t have a clue of how a paver even works. As I see it, it's not there fault, it's there employers fault for not training them, or ensuring they were trained. We would have a lot smoother roads if those involved were better trained.

Anyways, that’s my two cents.Good luck.

 

[LCruiser]

Moogbro -

Good explanation. My question to you is why do you say:
"If your surface you were paving over were really rough, the best you could hope for would be to have it about 30% as rough. This general rule is why legitimate specifications limit the use of a profilograph on projects where only two lifts using automatics can be placed,not to mention require the use of a paver with automatics and a minimum ski length."

I don't understand why that would be, unless you mean "*at least* two lifts." Still, you are specifying methods, which don't mix with performance specs such as a profilograph or IRI.

 

[LCruiser]

I forgot to mention. A big indication of a problem with paving is a paving machine that's not moving... The feed needs to be continuous and consistent.

 

[moogbro]

If I understand your question correctly, why specify a method (a fully automatic paver) when you have an end result spec (a profilograph or straightedge spec)? The answer is because you want a finished product which meets your ride tolerance and with a minimal amount of grinds to get it. Grinding not only looks bad, it also affects the surface of the roadways ability to repel water. Grinding removes tops of aggregates which makes them more likely to pop out, and also has a tendency to break the joints between the aggregates and the fines, making it more likely to degrade due to water intrusion.

Requiring the use of a fully automatic paver will give you the best chance of getting what you want, a smoother finished product. A contractor paving with automatic screed controls can still can still screw up a job, you just want to decrease the chance of it happening. If you don’t specify automatics, then end up with a contractor without automatic screed control, you will likely have more bumps that will need to be ground off. Automatic controls I believe cost about $20,000. Contractors aren’t stupid, if it’s cheaper to grind a few spots, why would they invest the $’s for automatic screed control. Contractors who don’t have automatic screed controls need to stick with paving projects where ride isn’t that important (parking lots, driveways, etc.).

In the above thread, I confused the issue by trying to add a little humor stating 30% as rough (i.e 100-70=30).

Additionally I tried to explain the 70 to 75% max improvement for one lift rule, and the 95% improvement for two lifts. I am not sure if I explained it clearly enough and will revise the paragraph to hopefully make it more clear (added text in caps):

One additional point to be made is that assuming that the screed spreads the material across a perfect plane, what happens after the rollers hit it. As the volume before compaction is 20% to 25% more than after the rollers hit it, thicker areas will compact more than thinner. THE RESULTING SURFACE AFTER COMPACTION WILL BE A SURFACE WHICH IS NO LONGER WITHIN A PERFECT PLANE. THE SURFACE OF THE THICKER SECTIONS WILL BE LOWER THAN THE SURFACE OF THE THINNER DUE TO THE COMPACTION (MORE MATERIAL TO COMPRESS IN THE THICKER SECTIONS) A general rule of thumb (PROVEN BY INDUSTRY TESTS) is the more lifts you can place, the smoother your ride will be. BECAUSE OF THE ABOVE, ONLY 70 TO 75% IMPROVEMENT IN RIDE IS POSSIBLE IF YOU LIMIT A PROJECT TO ONE LIFT, AND LIKEWISE ABOUT 95% IMPROVEMENT IN RIDE IS POSSIBLE IF YOU CAN PLACE TWO LIFTS. THIS ASSUMES THE PAVING OPERATION IS RAN PERFECTLY (i.e pulls a perfect plane).

The above is somewhat simplified because it does assume a the screed would immediately follow a tow point adjustment to end up with a plane parallel to the tow point path. In reality once the towpoint is raised the screed needs to travel 6 to arm lengths before it falls back into equilibrium (i.e following a parallel path to the tow point line), it's pretty darn close after traveling 3 tow arm lengths.

 

[ACtrafficengr]

"THE SURFACE OF THE THICKER SECTIONS WILL BE LOWER THAN THE SURFACE OF THE THINNER DUE TO THE COMPACTION (MORE MATERIAL TO COMPRESS IN THE THICKER SECTIONS) "

Which is why some agencies specify a shim or truing and leveling (T&L) course to fill ruts and deeper depressions before overlaying an existing surface. Although I understand some NY State DOT districts now grind off the top course prior to overlaying, because multiple layers of top course can be more vulnerable to rutting under truck loads.

"...students of traffic are beginning to realize the false economy of mechanically controlled traffic, and hand work by trained officers will again prevail." - Wm. Phelps Eno, ca. 1928

"I'm searching for the questions, so my answers will make sense." - Stephen Brust

 

[moogbro]

ACTrafficengr

Exactly, if you have deep ruts, fill them or grind down to a flat surface first.

Also, worth mentioning, which alot of designers don't fully understand. Some grinders (rotomills) also have automatic grade controls. They can be dialed in to grind off material at a constant cross slope and/or grind to a depth based on a stringline, (not to mention just setting down and grinding a constant depth into the existing pavement).

Sorry for being so long winded in the above threads. After I found this site, I have learned alot about other aspects of engineering I don't have alot of experience in, so figured I might as well try to teach others about what I do have alot of experience in. For young or inexperienced engineers working on paving projects, unless you attend a class on paving, you will likely not learn it on the job because contractors will typically not spend any time explaining it.

 

[zabrab]

Assuming the paving operations did not occur over a base with abrupt vertical transitions, vertical differences due to compaction of areas of deeper lifts should be somewhat gradual and would have more of an “undulating” ride, something that would be noticeable on a high-speed roadway.

“Bumpy” sounds as if there are discrete locations with some type of vertical discontinuity. That the roughness was noticeable just after completion of the project, I tend to agree with the other posters who attribute the roughness to inconsistent operation of the paver (paving stopped because of material delays, running the hopper out, operator bumps trucks, etc.) or problems (dips, ruts, depressions) in the underlying base that were not fully corrected. If it was a compaction issue, I would think you would see areas of rutting or raveling that you could easily identify.

All that being said, 2.5” does seem like large lift for a surface course. A max lift for a 9.5mm mix would be something on the order of 1.5”. I would have to check, but a 2.5” lift might be right at the max for a 12.5mm mix. It could be corrugation and shoving caused by traffic action combined with low stiffness in the surface course or poor bonding between the pavement layers. Although this would generally be localized at intersections, it can occur at various locations including some that are remote from the intersections if there are long queues as it occurs at points where traffic starts and stops.