The Great Inkwell Experiment

[amyx231]

I think "A" really supports my observation that my modern Montblanc hinged lid gravity closure inkwell was very poor at preserving ink. I may try to develop some sort of seal for it as It is quite nice to look at and would be even more lovely with some Kon-Peki or Asa-Gao in it...but not at 35% loss in two weeks (that translates to $5/week loss )

 

distilled water. just add water :ltcapd: .

[WOBentley]

I think "A" really supports my observation that my modern Montblanc hinged lid gravity closure inkwell was very poor at preserving ink. I may try to develop some sort of seal for it as It is quite nice to look at and would be even more lovely with some Kon-Peki or Asa-Gao in it...but not at 35% loss in two weeks (that translates to $5/week loss )

 

distilled water. just add water :ltcapd: .

Understood, -and that is what I have done, but it is a hassle and at the cost of those Montblanc inkwells, you think you would get a better seal!

[amyx231]

rubber ring/washer glued to the top? or a few rings set around the inkwell glass bottom opening, so the top closes on rubber?

[D Armstrong]

Week three results are in!

 

http://www.restorersart.com/images/blog/inkwell_experiment/Inkwell%20Experiment%20Results%20Chart-Week%20Three.gif

 

 

And it seems that I'm finally getting the hang of Excel (or, more specifically, how to make readable charts using Excel.)

 

As expected, the brass inkwell went dry last Saturday. Interestingly, even the modern, very effective, ink bottle is exhibiting some loss over time.

 

And it seems that the time-honoured solutions of using natural products such as cork (the stopper for F) and wood (the case for D) are very effective. It reminds me of the Japanese chests, made of paulonia wood (the same wood as many traditionally-made Japanese kitchen knives). One significant characteristic of paulonia is that it absorbs huge amounts of water, very quickly. The reports are that, when caught in a flood, the drawers of these chests swell shut almost instantly, perfectly preserving the clothing inside them. It made me wonder if the wood of inkwell D is simply swelling shut enough to limit air exchange, and thus evaporation.

 

In any case, one more week to go, and then on to conclusions. Tune in next Friday...

[WOBentley]

fascininating and somewhat surprising results...

I can't tell e and f apart, but clearly they seem to be doing well if I am reading this correctly!

Thanks for the work!

Dave

[celesul]

Are you also keeping track of the volume in mL, or some such objective measurement? If the amount of liquid in each inkwell is very different, then it isn't a fair test. On the other hand, if the volume is generally similar, then the absolute measurement is relatively inconsequential. Would you perhaps be able to tell us at least the initial absolute measurement to give us a better sense of what the data means? I'm simply rather curious. Your experiment is very awesome!

[beak]

Very interesting experiment - and well conducted, it seems. Thanks.

 

But I'm having a lot of trouble with C - doing worse than the control!

 

Have you any idea why this is so; it seems so unlikely, given that C has at least some protection from evapouration?

[D Armstrong]

Are you also keeping track of the volume in mL, or some such objective measurement? If the amount of liquid in each inkwell is very different, then it isn't a fair test. On the other hand, if the volume is generally similar, then the absolute measurement is relatively inconsequential. Would you perhaps be able to tell us at least the initial absolute measurement to give us a better sense of what the data means? I'm simply rather curious. Your experiment is very awesome!

 

If you take a look at the PDF it contains all of the raw data in a spreadsheet. The total amount of liquid in each varies widely, depending on the volume of the container. This experiment is based on relative amounts, rather than absolute, as we are testing the relative air-tightness of each. If we had the same amount in each, then the larger air mass in some would skew the evaporation rates (as there would be more space in to which the water vapor could escape.)

[D Armstrong]

Very interesting experiment - and well conducted, it seems. Thanks.

 

But I'm having a lot of trouble with C - doing worse than the control!

 

Have you any idea why this is so; it seems so unlikely, given that C has at least some protection from evapouration?

 

I've been banging my head on this one for three weeks! At first I thought the thing must leak, but no, it's sound. Small animals sneaking in every night and feeding off the water? Aliens? Can't find anything.

 

So, as mentioned above, I think I'll run another experiment later, using inkwells of similar design (brass cases with inserts and loose caps), just to see what other variables may be in play.

[wheezur]

Very interesting experiment - and well conducted, it seems. Thanks.

 

But I'm having a lot of trouble with C - doing worse than the control!

 

Have you any idea why this is so; it seems so unlikely, given that C has at least some protection from evapouration?

 

I've been banging my head on this one for three weeks! At first I thought the thing must leak, but no, it's sound. Small animals sneaking in every night and feeding off the water? Aliens? Can't find anything.

 

So, as mentioned above, I think I'll run another experiment later, using inkwells of similar design (brass cases with inserts and loose caps), just to see what other variables may be in play.

 

I'm not at all surprised that some ink wells lost water by evaporation faster than your "open" control.

 

Your "open" control is a tall graduated cylinder. The rate of loss from a cylinder is inversely proportional to the distance between the liquid level and the top of the container. In your case the distance between the top of the liquid and the top of the cylinder appears to be pretty large: maybe 6 inches (I can't really tell from the pics)? In an open ink well or open ink bottle, the distance would be something like half an inch, or maybe an inch. Yes, it changes with time, but on average will be something like that.

 

If you had cut your graduated cylinder so that it had similar dimensions to an ink well or ink bottle, you would get an evaporation rate that is something on the order of 6 to 12 times faster. My estimates of the dimensions are probably off, but it's pretty clear that the multiplier will be large and the evaporation rate would be much faster than for even the poorest ink well.

 

You have demonstrated nicely that evaporation from a tall cylinder is surprisingly slow, making it better than a short ink well with a poorly fitting lid - at least from an evaporation standpoint. I suspect that it would be difficult to fill a pen from the graduated cylinder - so it would not be a practical ink well.

 

Edited to fix typos.

Edited March 14, 2011 by wheezur

[Sleepy]

Very interesting experiment - and well conducted, it seems. Thanks.

 

But I'm having a lot of trouble with C - doing worse than the control!

 

Have you any idea why this is so; it seems so unlikely, given that C has at least some protection from evapouration?

 

I've been banging my head on this one for three weeks! At first I thought the thing must leak, but no, it's sound. Small animals sneaking in every night and feeding off the water? Aliens? Can't find anything.

 

So, as mentioned above, I think I'll run another experiment later, using inkwells of similar design (brass cases with inserts and loose caps), just to see what other variables may be in play.

 

I'm not at all surprised that some ink wells lost water by evaporation faster than your "open" control.

 

Your "open" control is a tall graduated cylinder. The rate of loss from a cylinder is inversely proportional to the distance between the liquid level and the top of the container. In your case the distance between the top of the liquid and the top of the cylinder appears to be pretty large: maybe 6 inches (I can't really tell from the pics)? In an open ink well or open ink bottle, the distance would be something like half an inch, or maybe an inch. Yes, it changes with time, but on average will be something like that.

 

If you had cut your graduated cylinder so that it had similar dimensions to an ink well or ink bottle, you would get an evaporation rate that is something on the order of 6 to 12 times faster. My estimates of the dimensions are probably off, but it's pretty clear that the multiplier will be large and the evaporation rate would be much faster than for even the poorest ink well.

 

You have demonstrated nicely that evaporation from a tall cylinder is surprisingly slow, making it better than a short ink well with a poorly fitting lid - at least from an evaporation standpoint. I suspect that it would be difficult to fill a pen from the graduated cylinder - so it would not be a practical ink well.

 

Edited to fix typos.

 

Thanks for getting there before me, saved me typing something similar. Wheezur is correct, your control is not a very good control. But to be honest, no one is going to be using an open cylinder for an inkwell anyway, so we're not all that interested in the performance of this control.

[D Armstrong]

 

Your "open" control is a tall graduated cylinder. The rate of loss from a cylinder is inversely proportional to the distance between the liquid level and the top of the container. In your case the distance between the top of the liquid and the top of the cylinder appears to be pretty large: maybe 6 inches (I can't really tell from the pics)? In an open ink well or open ink bottle, the distance would be something like half an inch, or maybe an inch. Yes, it changes with time, but on average will be something like that.

 

If you had cut your graduated cylinder so that it had similar dimensions to an ink well or ink bottle, you would get an evaporation rate that is something on the order of 6 to 12 times faster. My estimates of the dimensions are probably off, but it's pretty clear that the multiplier will be large and the evaporation rate would be much faster than for even the poorest ink well.

 

You have demonstrated nicely that evaporation from a tall cylinder is surprisingly slow, making it better than a short ink well with a poorly fitting lid - at least from an evaporation standpoint.

 

 

The distance from the initial water level in the cylinder was about 21mm, and the correlating distance for the brass/insert inkwell was about 14mm (obviously estimated, as I didn't think to measure at the time.) It seems to me that such a multiplier shouldn't make that much of a difference. Although I could be wrong.

 

And, of course, most inkwells are just fancied-up cylinders, of various dimensions.

 

 

I suspect that it would be difficult to fill a pen from the graduated cylinder - so it would not be a practical ink well.

 

 

Hmmm. It's actually just a taller version of what Brian Goulet uses to send out his "ink drop" samples, and is what I use in the shop normally to soak and flush pens I'm working on. Works like a charm.

 

That being said, I did choose a larger container because I wasn't sure just how much volume I would lose out of the open control. And I made sure to choose one which had a similar diameter opening to an average inkwell, so as to avoid the larger-surface-area-equals-faster-evaporation-rate complication.

 

Also, the purpose of controls in this case is to establish the extreme benchmarks, to put the other results in an understandable context. It's just that the extreme loser ended up being not so bad after all. Go figure.

 

There might be something highly philosophical in there somewhere...

 

Thanks for the input, it keeps me on my toes.

[wheezur]

 

Your "open" control is a tall graduated cylinder. The rate of loss from a cylinder is inversely proportional to the distance between the liquid level and the top of the container. In your case the distance between the top of the liquid and the top of the cylinder appears to be pretty large: maybe 6 inches (I can't really tell from the pics)? In an open ink well or open ink bottle, the distance would be something like half an inch, or maybe an inch. Yes, it changes with time, but on average will be something like that.

 

If you had cut your graduated cylinder so that it had similar dimensions to an ink well or ink bottle, you would get an evaporation rate that is something on the order of 6 to 12 times faster. My estimates of the dimensions are probably off, but it's pretty clear that the multiplier will be large and the evaporation rate would be much faster than for even the poorest ink well.

 

You have demonstrated nicely that evaporation from a tall cylinder is surprisingly slow, making it better than a short ink well with a poorly fitting lid - at least from an evaporation standpoint.

 

 

The distance from the initial water level in the cylinder was about 21mm, and the correlating distance for the brass/insert inkwell was about 14mm (obviously estimated, as I didn't think to measure at the time.) It seems to me that such a multiplier shouldn't make that much of a difference. Although I could be wrong.

 

And, of course, most inkwells are just fancied-up cylinders, of various dimensions.

 

 

I suspect that it would be difficult to fill a pen from the graduated cylinder - so it would not be a practical ink well.

 

 

Hmmm. It's actually just a taller version of what Brian Goulet uses to send out his "ink drop" samples, and is what I use in the shop normally to soak and flush pens I'm working on. Works like a charm.

 

That being said, I did choose a larger container because I wasn't sure just how much volume I would lose out of the open control. And I made sure to choose one which had a similar diameter opening to an average inkwell, so as to avoid the larger-surface-area-equals-faster-evaporation-rate complication.

 

Also, the purpose of controls in this case is to establish the extreme benchmarks, to put the other results in an understandable context. It's just that the extreme loser ended up being not so bad after all. Go figure.

 

There might be something highly philosophical in there somewhere...

 

Thanks for the input, it keeps me on my toes.

 

 

I'm surprised (obviously) by the dimensions you give - I thought that the graduated cylinder was much taller. With the values you give, the multiplier would only be about 1.5 and that would not be enough to make the evaporation rate from the open cylinder greater than the evaporation rate from container C. So, the mystery remains . . .

[amyx231]

perhaps you can try it with the top off, for C? maybe the top is actually sucking up water? As in, getting soggy or something?

[D Armstrong]

perhaps you can try it with the top off, for C? maybe the top is actually sucking up water? As in, getting soggy or something?

 

Mmmm, no. It's made of hammered brass, which does not absorb water. In fact, no part of this one absorbs water, nor does it leak (I checked.)

[amyx231]

perhaps you can try it with the top off, for C? maybe the top is actually sucking up water? As in, getting soggy or something?

 

Mmmm, no. It's made of hammered brass, which does not absorb water. In fact, no part of this one absorbs water, nor does it leak (I checked.)

 

k. i thought there might be glue or something left, from maybe fabric long-ago.

[D Armstrong]

And the final results are in!

 

http://www.restorersart.com/images/blog/inkwell_experiment/Inkwell%20Evaporation%20Rates-final.gif

 

 

A) glass well with a simple gravity-closing hinged lid;

B) glass well with a silver snap-close hinged lid;

C) metal well with glass insert and a hinged outer cover (thus with space between the insert and the cover);

D) wooden well with glass insert and an unhinged cloth-lined wooden lid;

E) glass writing-box bottle, with threaded brass lid;

F) pottery ink bottle, with a new, replacement, cork;

G) (control) plastic graduated cylinder, open;

H) (control) modern glass ink bottle with plastic lid and synthetic lid liner (J. Herbin)

 

 

The full raw data is here.

 

And as for observations and conclusions? I'm working on it, and the final documentation.

 

Any thoughts?

Edited March 21, 2011 by D Armstrong

[watch_art]

the graphs would be easier to read if the letters were arranged in the same order as the colored lines, instead of alphabetically. can you change that?

[amyx231]

i can handle the lines being alphabetical. but can you please write the names/tags of the bottles instead of letters alone? I keep having to scroll back up....

[D Armstrong]

I've edited the post to include the key, so you don't have to scroll.

 

As for tweaking the colors or order, that would require someone whose powers over Excel are great, indeed...