![Sample our Engineering & Technology journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5933/TOC-Subject-Sample-2021-Engineering-Tech.jpg"})
Abstract
Asphalt foaming process, as one of the major warm mix asphalt (WMA) technologies, can significantly increase the volume of asphalt binder with large surface area in the unit volume and thus generate a strong coating with high shear strength of the mix and improved workability. It allows lower production and construction temperatures and less greenhouse gas and other emissions. Foamed asphalt has generally been achieved by introducing water as a foaming agent into hot asphalt flow before mixing with aggregate at a certain temperature. The water will evaporate and make the asphalt’s volume expanded and its viscosity reduced. However, both the formation and decay of the foamed asphalt binder is a highly thermodynamic process, which makes the characterisation of the foamed asphalt binder extremely difficult. A high fidelity model to simulate the foaming process will provide a powerful tool to conduct virtual experiments of WMA production and optimise the asphalt foaming process and WMA design. To this end, a numerical model using the smooth particle hydrodynamics is developed to simulate the asphalt foaming processes. A self-developed nozzle-based foamer was used to generate foamed asphalt binder at different water contents. Three primary parameters, expansion ratio, half-life and foam index that are widely applied to evaluate the foaming characteristics of foamed asphalt, have been studied. It was found that simulation results agree well with the experiments. Parametric studies were further conducted by using the numerical model to evaluate the effects of environmental controlling parameters on the foaming characteristics of the foamed asphalt binder.
Acknowledgements
The authors thank Dr Liming Li and Ms Gisele Passalacqua for their contributions on the laboratory testing. The authors also thank the Axeon Specialty Products for their provided asphalt binder used in this study.
Discussion
BOB KLUTTZ: I like the modeling very much. I think we need to do more work like that. I think it’s going to be very helpful. Two comments. I think it would be very interesting, especially considering the paper that we saw yesterday, to look at some different bitumens because I think the results that you get may be quite different with that, and I think it would be a very interesting opportunity to validate your model if you can simulate the different behavior with different bitumens.
HUIMING YIN: Yes, that’s true.
BOB KLUTTZ: Okay. Second comment. I hope you're not actually encouraging people to physically use ethanol to foam bitumen. Ethanol is flammable. I cannot imagine doing this in any kind of commercial process that will not result in an explosion. I strongly discourage laboratory trials of that.
HUIMING YIN: Yeah, I think that’s a good comment, but our collaborative team of Michigan Tech and Columbia University have already done a lot of study about it, like flaming point, to avoid safety problems. In this way if we do it with ethanol, we only need 1 to 2% of ethanol compared with binder. If you put into the counter-flow drum mixer, as we talked with the people from Astec, it is pretty safe and can be integrated into their existing production process.
BOB KLUTTZ: It has nothing to do with the amount of ethanol in the bitumen. It has to do with the amount of ethanol and air in the gas phase. And ethanol in air is explosive at 3% to 19% by volume.
HUIMING YIN: Actually, if we integrate the eFOAM process into the existing mixer, we can reuse the evaporated ethanol burned in the flaming side with gas together. For a certain kind of counter-flow mixer, you can still do it. But, yeah, we’re still considering how to make it work in an actual field project.
DAVE NEWCOMB: Very good paper on the modeling aspects of foaming, and I was very encouraged to see how well tracked your model.
HUIMING YIN: Thank you.
DAVE NEWCOMB: But there are a couple of issues and one that Bob Kluttz just brought up with regards to the source of the asphalt. We in NCHRP 9-53 tested several different sources of asphalt, and they range all over the board in terms of their ability to foam and at what water content they reduced to kind of an optimum foam. Which kind of brings me to my next comment, and that is that all this is well and good to measure in the laboratory and to model, but what we found was that the real difference with respect to mixture behaviour in terms of workability and coatability came from the water content more so than any over parameters that we measured such as ER, HL, foamability, and things like that. So, if you're going to go forward with this research, I strongly suggest that you incorporate some mixture behaviors in it, and I think you’ll find that that kind of trumps a lot of the foaming characteristics.
HUIMING YIN: Thank you. I think those are very good comments. After we end this project, we have kind of a fundamental research project from the National Science Foundation (NSF). For Columbia’s side, we’re working on binder. As for the mix design, it is done by the team from Michigan Tech. They're still working on that part. Their work may address the mix and long-term performance. That’s very important in actual application. We still need to get our parameters for each type of asphalt from lab tests, which means we reduce our lab tests by simulation, but do not replace our lab tests.
MIKE HEITZMAN: I’ve been working with foam for better than ten years, and I’m curious as to how you came up with this setup of your equipment because it doesn’t represent the typical type of foaming system that’s used in the industry.
HUIMING YIN: Thank you. This research is from National Science Foundation. Our collaborator purchased one foamer from Instron. That’s designed for water; however, we want to study for ethanol. We found every time we put ethanol inside, the foaming machine got clogged. So, we couldn’t do the foaming tests at all with ethanol. Then we developed this lab version so we can do both water and ethanol. So that’s the background.
MIKE HEITZMAN: In one of your early slides you showed a foaming chamber that’s all contained in a very small nozzle. And I don’t think that’s what you have here, so you don’t really have a value of the pressure that’s being generated during that expansion timeframe.
HUIMING YIN: Yeah, so actually we have pressure pipelines to control pressure separately. The first one is the asphalt pressure meter to show how much the pressure is in the asphalt flow. The second one is the water pressure. We also control the air pressure with this one in the chamber, so that’s three pressures. We could pump air inside to keep the chamber pressure.
MIKE HEITZMAN: I just suggest that as you move forward, look at what the industry standard is for foaming and try to work with that device.
HUIMING YIN: Thank you. We purchased the one foamer, but it doesn’t work very well. This one is a very simple one. We can definitely refine with an industry collaborator too.
MATTHEW CORRIGAN: We’ve been working with the industry to collect warm-mix asphalt usage data in the United States now since 2009. And from that data clearly it shows that foaming using water injection dominates the industry as far as usage. And so, when you tie that into the work that Dr. Newcomb did in the NCHRP 9-53 project, one of the things that he discovered out on actual field projects is quite often the contractor had an improper water injection ratio to where they didn’t get optimum foaming characteristics to get that coating and workability. And I’m looking at this maybe from a different viewpoint that you’ve got simulation so that if you can characterize several different materials – probably not all but a lot of different asphalt sources, different types – and be able to come up with a performance envelope of characteristics that contractors could use to make sure that they're in some range of operation that would get them close. Because there’s going to be some reluctance to buy any type of equipment like this to do that kind of evaluation to optimize the volume of their foam or the coatability and workability. But if we could use the simulation as a practical tool to get them closer so that when we’re out in the field and they don’t have their process optimized and they're not getting adequate coating or adequate workability, we can give them some guidance without having to go through actual testing evaluation but we could use your simulation potentially to get them within some boundary condition. So, are you guys exploring maybe the potential of that kind of guidance?
HUIMING YIN: I think that’s a very good suggestion to develop kind of a workable range, for example, for temperature, water content, and pressure. I think that’s a good suggestion. Unfortunately, we haven’t started this type of work yet. We would be very happy to collaborate with the industry or the agencies to develop such kind of specifications.
MATTHEW CORRIGAN: Thank you.
HUIMING YIN: Thank you.
YOGESH KUMBARGERI: I have one question regarding the modeling of the _____. With different kind of binders, like rubberized or polymerized if you use it, then how do you ____ the effect of that in our ____ model?
HUIMING YIN: Yes. Thank you. We have here, that’s the parameter for the physical property, then we can measure that directly from materials. For the simulation parameter, we need to do one test at least. Maybe you can do a couple tests to get a better average to be more representative. For example, the collapse rate may change with different types of binders. We needed to do comparison with experiments to determine these parameters. The bubble may collapse very quickly for a certain kind of binder. For some other types of binder, the rate may be quite small. That is dependent on the material type. After all, the collapse rate is also related to the water content. That’s a unique one. But for the specific heat of for asphalt, typically changing a very narrow range, we can make it kind of an input parameter without too much additional effort. But for the collapse rate, that’s really related to the surface property of the asphalt. So, then it can change significantly.
AKSEL SEITILARI: My question was during the simulation; do you consider the bubble size? Do you have a threshold for the diameter of a bubble? And how do you define that?
HUIMING YIN: That’s a very good question. We didn’t really get into physical bubble size here because we focus on the ER. The bubble size in the current work really depends on our cutoff size. The cutoff size in our simulation is given at 1 cm. This is how we determined the bubble size. I can make the cutoff smaller then to get smaller size particles. That’s a kind of simulation parameter. It changes with the physics of materials.
AKSEL SEITILARI: Thank you very much.
HUIMING YIN: Yes. Thank you.
AUTHOR'S CLOSURE
This paper is the first step to use a computer to conduct virtual experiments for the asphalt foaming process, which is very hard to directly measure inside of the foaming chamber. With a few parameters determined by lab tests, it reproduces the complex foaming process on the computer. It can be used by the industry, government agencies, and academia for design, development and optimization of asphalt materials and other composites. We look forward to seeing more application of this tool in the design and development of novel asphalt materials.