Further research concerning the dynamic angle of repose

Posted on by Damian

I’ve taken a look at a couple of papers both from the journal Pharmaceutical Research. Neither paper seemed concerned with the desegregation of the powders they were analyzing. The paper by Concessio et. al. did mention making mixtures with some of their powders (involving micronized albuterol sulfate) which they mentioned having mixed at 20rpm, but they didn’t mention whether this speed was determined to be the best or if it was arbitrarily chosen. To calculate the dynamic angle of repose they had a similar setup with a rotating drum rotating at .5 rpm, but once again it wasn’t mentioned why this speed was chosen. It seems that they captured video and then took measurements every 100ms to calculate the dynamic angle of repose of the powder. Another thing that caught my interest was that moisture contend did not seem to affect the static angle of repose as I would have assumed, perhaps surface morphology is a larger contributor to this factor. The second paper (by Lavoie et. al.) talked about the static angle of repose which was measured by letting powder drop from a funnel 5cm from the base until a stable cone was formed by the powder. Interestingly enough they also measured flowability and chose to measure this over wide set of speeds, unfortunately they didn’t gather any information regarding the dynmic angle of repose which would have been interesting to compare. Their static angle of repose method seemed to generate multiple angles of repose for a few powders as presented by their data. I wonder if this would also be true for the dynamic angle of repose except with a speed/rotation dependency. They defined some methods to determining cohesion and flowability indices, interestingly enough I didn’t see a significant correlation between these in the powders that they used (this is just a qualitative observation done by looking at some of their graphs/data presented). I’ve skimmed a few other papers but they weren’t as interesting as these two and none seemed to have any mention of rotation speed as a factor for the dynamic angle of repose in pharmaceutical powders. The best approach I can think of is taking multiple readings of the dynamic angle of repose at different speeds, and then performing a statistical test to determine whether any differences are significant.

Paper 1: DOI: 10.1023/A:1018865717374

Paper 2: DOI: 10.1023/A:1016125420577

Thoughts regarding the revolving powder analyzer

Posted on by Damian

I read a paper recently dealing with what I think was a binary particle mixture. The paper explored the segregation of said mixture due to the effects of being in a rotating drum. While trying to take data with the revolving machine in the lab I wondered if perhaps the lactose mixtures we were using might be affected by rotational speed, number of rotations, and the volume used. This paper seemed to provide some insight into the mechanics these kinds of things, but it is clearly stated that the specifics are intrinsic to the material in question. That being the case I doubt I can make any solid conclusions regarding the lactose we have been using in the lab. Segregation of the particle mixture took longer when the volume in the drum was larger, and seemed to follow some sort of power law. Another non-intuitive finding in the paper was that as the ratio of the particles diameter (d1/d2) decreased from one, it took more revolutions to achieve maximum segregation. This seemed to have more of linear relationship. It was also mentioned in the paper that at higher speeds the particles pouring down (avalanching) over the open surface became significant enough so as to inhibit segregation at least partially. I’m not sure how these effects would manifest themselves in a microscopic size range, perhaps these only apply to macroscopic particles as it seems that’s what the paper was dealing with. If the mechanics were to be similar I feel that since we are trying to obtain the avalanche angle of the powder as a homogeneous mixture, then perhaps we should use relatively high volumes (i.e. almost fill the small drum) and use high rotational speeds. Perhaps a test could be performed to see if these two parameters matter and if so how they affect the avalanche angle.

Paper referenced: Eskin, D., Kalman,  H. (2000). A numerical parametric study of size segregation in a rotating drum. Chemical Engineering and Processing, 39, 539-545.

Regarding SuperTab30GR, Rotating Powder Analyzer, and some new labeling methods for the powders

Posted on by Damian

Today I once again tried to get SuperTab30GR to produce some usable results, but it seems that the powder is displaying some annoying/unique properties. The sloping it produces isn’t exaggerated enough to one specific side so that it can be clearly visualized from an angle. It’s irregular, bumpy, surface that significantly deviates from the shadow it produces at any angle seen is hard to deal with. I loosened the arms holding the cylinder, just enough to rotate it so it’s position wouldn’t be affected, but nothing really struck me as a good represented of what the powder was doing. I also feel that if the angle at which the cylinder is placed is exaggerated too much, then the mechanics of the system change and thus the data itself becomes different from the data that’s already been collected. Some things that might work but would also be a change in the variables thus far are: tapping from a different height, even out the surface between taps somehow, the thought also occurred to me that perhaps compaction rate (versus time) isn’t affected as long as one takes only a section of the powder that seems to constantly decrease up until it reaches the maximum compaction. That is the kinetics are represented by a section of the powder, and the whole surface need not necessarily be sampled. I’ve also starred the powders that have had all 3 sets of data taken, so as to organize and facilitate keeping track of these. Below is a video showing what I talked about regarding SuperTab30GR.

I also got around to taking some screenshots of the software that’s associated with the rotating powder analyzer, and how it’s supposed to work. There are a lot of pictures that are going to follow from this point on. It’s broken down into simple steps, not meant to be patronizing only easy to follow.

First you create a new analysis template, in our case we want to take data of the powders “Flowability.”

Then there are several different settings that you can mess around with, some of which I’m not quite sure what settings would be best. I’ll point those out when the picture comes up. I think it was already agreed that we’d be using the small drum but I’m not sure if the data should be normalized for this. With regards to requiring the sample weight, I think this is best so one can’t start without having weighed the sample so as to avoid forgetting to do so.

This next picture is an example of some settings I’m unsure about. I think having an imaging rate of 30 FPS (the max) is best, but I’m not sure about the rotation rate, avalanche threshold, the number of avalanches observed, or how the angle should be calculated.

Another picture below of some settings I’m unsure about.

Again, this seems less important but the settings might still have an effect on the data that’s collected.

Not really sure if all parameters should be displayed or if we can exclude some since they won’t be relevant.

Here you select where to save your results, given that “Save Results” is checked. You can also choose to output a text file with the data and a graph. This seems to create an Excel file but it’s poorly organized so I’m not sure how much use this option really is.

Some printing settings that I don’t think are that important.

Once you are done, you save it. As you can see I already had one made named “flowbility-smallDrum.fam”

To make a mask one simply selects “New Mask” from the mask menu item.

You’ll get this screen. You take a picture (with the drum of your choosing in the machine) by clicking the camera icon.

There was no drum at the time so the white area is relatively big compared to when a drum is present. Then you mask the area you want to be the focus point using the red button with the letter “m” inside.

It’ll turn out like this.

You then save your mask file. I already had one for the small drum named “smallDrum.ama”

To start an analysis of a powder you click the circle icon that has it’s bottom half shaded a dark blue/black.

Then this screen comes up, you select the test method that you previously created, select the mask you made as well, name your sample, place any comments, and finally input the sample weight which I’ve set to required in the test method I created (i.e. you can’t start the analysis without inputting the weight).

Rotating Powder Analyzer

Posted on by Damian

Today I got myself situated with the rotating powder analyzer. I got a better feel for how to work it and also how to make a mask for a given drum. The process is relatively simple unfortunately I couldn’t really get any screens of it since the computer does not have an internet connection and I didn’t have a USB on me at the time. The analysis that we are performing on the powders is the “Flowbility” one which gives detailed information about the avalanche angle of the powder (i.e. the angle of repose). There are several parameters that I’m not sure about, that is I don’t know what settings should be used on these. For example there is an RPM setting, a setting affecting where the measurements are taken, and a few more settings along with some output parameters of which I don’t know are needed. I suppose I could set the output to display as much information as possible but that still leaves the other settings that affect the actual data taking process unresolved. Learning how to work the machine after these parameters are defined isn’t hard as it is automated process from there on. 

Posted on by Damian

Just some samples from a few trials I did yesterday and today. These show some piles formed by the fine powdered lactose. I tried to replicate an effect Dr. Smyth mentioned about rippling when the powder is let fall down. Although I could kind of see it start to happen in the second to last photo, I really need to make something that can hold the funnel and has a removable piece to let the powder out. I’ve some ideas which I’ll get to working on soon.

Posted on by Damian

Here are some trial runs of the angle of repose lactose project. I did some trials with the lactose that is bigger in size and it had a hard time flowing out of the cylinder and didn’t form mounds as well. I also noticed that the fine powder lactose flows out better when it’s “fluffed.” The last picture shows the mound that formed from the fine powder lactose when it was compacted (i.e. tapped). The mound looks smaller because only about half the lactose came out and it came out in “chunks” (i.e. it didn’t flow out smoothly).

Posted on by Damian

I made a device about a week ago to hold a cylinder and allow me to tilt it while holding it at a constant height. This shows the previous form of it, as I had to end up making the holding block the piece that rotates directly instead of having it attached to something that rotates. This allowed for the location of the falling lactose to be kept relatively constant which was not the case for the form seen in these pictures (the position kept varying as a function of the angle the cylinder was in).