I am in the market for a winter bag. Obviously, I consulted this great forum and adjacent articles but it eventually left me in the blue with regard to the choice between:

- small (read: lighter) shell & baffles with a significant overfill (e.g. 100% overfill, thus slightly compressed down)
OR
- bigger shell & baffles with no or limited overfill (thus resulting in normal loft).

‘Über’ insulation expert Richard Nisley seems to advocate overfill. In a discussion (http://www.backpackinglight.com/cgi-bin/backpackinglight/forums/thread_display.html?forum_thread_id=12505&disable_pagination=1), he refers to two research results:

- Fig 9 of Application of Nanofiber Technology to Nonwoven Thermal Insulation" Phillip W. Gibson, Ph.D.1, Calvin Lee, Ph.D. 1, Frank Ko, Ph.D.2, Darrell Reneker, Ph.D.3 (see p.5 http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA481751)
- Richard’s own research whereby he overfills a bag up to 2,5x.

Firstly I looked at the graph of the Gibson & Lee paper. It shows the thermal conductivity of several insulation fiber on the basis of kg/m3, going from 0 kg/m3 up to aprox. 80 kg/m3, at least for down. Then I took my calculator and discovered that 850 CUIN (European measurement) equals 4912m3 per kg. Does that not render the graph only applicable for highly compressed down, since it does not clearly show the impact with extremely low kg/m3 . Or is my reasoning and/or calculation wrong?

Then, there are the – debatable – measurements of Jerry Adams (http://www.backpackinglight.com/cgi-bin/backpackinglight/insulation_measurement.html?forum_thread_id=65144). One of his setups was a piece of down overfilled by 100%. His measurements was an increase of clo but not even in the neighborhood of an linear increase.



In real world, the correlation might however perhaps be a bit stronger, a.o. because of head radiation & smaller shell (read: heat exchange surface). It would however surprise me if the a real world setup would result in an (almost) linear relationship.
On the other side is Richard’s own research, whereby he overfilled bag(s) up to 2,5. The clo-correlation between total bag weight and clo-increase was virtually 100%. It is possible to give plausible explanations for these measurements.
One thing I have not seen in the discussion boards is the impact of condensation on overfilled sleeping bags. A thesis might be that condensation will be worse as there is more mass per cubic inch.

Question: do I have to be talked out of significantly overfilling and rather go for a more conventional approach? I am a bit confused in this department (and the new sleeping bag will be a relatively big investment).

PS: apologies in advance for any potential spelling error(s). English is not my mother tongue

"debatable"

debatable? debatable???? (just kidding)

Richard's the expert and I really appreciate all the information he's given out, but I disagree with him a little on this one point

I measured 1.29 clo/oz.yd2 for normal loft, 0.96 for 100% overfill - so you're better off allowing the down to loft to it's maximum

Richard has a chart in http://www.backpackinglight.com/cgi-bin/backpackinglight/forums/thread_display.html?forum_thread_id=12505&startat=20 "the effect of down density, in a fixed 3.25" baffle size, on the insulation value of sleeping bags"

On the left side is 4.5 clo for 16 ounces of down = 0.28 clo/ounce
On the right side is 5.8 clo for 28 ounces of down = 0.21 clo/ounce
That's pretty consistent with my data

But, I think you can overstuff up to 2.5X and still get an increase of clo. Maybe above 2.5X you get no increase?

And if you're using American standard fill power, you want to over-stuff by 10% because the spec is less conservative. European standard you don't have to do this.

And maybe overstuff another 10% or 15% just to make sure all the baffles stay full and the down doesn't shift to one side. As Richard points out you get an increase of clo so it's not wasted.

So if your sleeping bag is 3 square yards = 3888 square inches, and you want 2 inches of loft, the volume is 7776 cubic inches, and you're using 850 fill American standard you want 7776 / 850 = 9.1 ounces of down - add 10% for american standard = 10.1 ounces - add 10% overfill to make sure baffles are full = 11.1 ounces.

Jerry--Correct me if I'm wrong, but it seems like your measurements are based on the weight of the down alone, but that's not the only factor contributing to weight. A larger bag, which does not compress the down, weighs somewhat more than just the added down itself. This does, of course, depend on design and material of the sleeping bag, so that's a huge factor.

I looked at Tim's Rev X bag for an example. It seems like for every 2.3 ounces of 850 fill down he adds, the weight increases by 2.5 oz. Not a huge difference, so if Jerri's measurements are correct, the loss of CLO from compression will make more of a difference. I'm sure it would be even more dramatic with a cuben quilt or some lighter fabric.

More generally, it was always an assumption of mine that the down will loft most ideally with extra space. That's how sleeping bags have always been designed, right? There must be a reason why.

My question is this: Most of the measurements referenced above use a very serious overfill, but what about a 50% overfill? Is the relationship necessarily linear, or might there be an ideal overfill level?

I was just talking weight of down

Linear? Look at that chart of Richard's. He did a linear regression producing that line, but you can see a slight curve to the actual data points. I think Richard talked about that.

Hmmm - thinking about it, it seems like the curve is going the wrong way - as you add more and more down it seems like at some point the clo would quit increasing??? - maybe that's just measurement noise.

In Richard's chart, if you assume 16 ounces is 0% overfill, then 24 ounces would be 50% overfill. 16 ounces = 4.5 clo = 0.28 clo/ounce. 24 ounces = 5.3 clo = 0.22 clo/ounce. You lose some efficiency but I think it would be hard to tell based on how warm you were sleeping in it and how tired you were carrying the extra weight.

At some point this becomes an engineer's obsession with optimization rather than anything that makes any practical difference

@jerry - if you look at that same thread just a bit below you will see Richard replied basically saying that when you divide TOTAL CLO b y TOTAL bag weight you get the same result over the whole range!!! which would indicate that there is almost NO LOSS of efficiency as far as CLO/oz over that range!! Note that the bag is THE SAME BAG - eg same baffle size and same weight....

Now putting aside stuff like measurement error etc (i assume richard is pretty aware of that methodology wise) and assuming richard's measured data is correct what I think the issue with your calc is that there could be some underlying process here (which i cannot say what it is) where by what you are trying to do (essentially subtract the CLO of the empty but fully lofted bag from the total clo to ostensibly ascertain the CLO or the down alone) is not scientifically justifiable. Now im not SURE of this of course, but maybe it has to do with the fact that usually when we add R values its when we put something ON TOP of something else and then its like resistors in series...but here you have something INSIDE something else and that is neither series nor parallel but maybe a more complex phenomena ...

all im saying is that although your MATH is correct the conclusions you draw from it are in contradiction to the totalclo/total weight measurements that richard quoted ....so there might be something wrong with your SEPARABILITY ASSUMPTION

@Wim - to give you my perspective of the answer (YMMV etc) - i assume what richard showed us is essentially correct AND as @NIck pointed out what you are comparing is a BIGGER HEAVIER bag with full loft vs SMALLER LIGHTER bag with compression so you need to account for the xtra weight of the empty bag BUT (!!!) I would not take it that far as 100% - remember that if the nominal compression is XX% then when your elbow/knee/head pushes against the bag or you layer in it there will locally be more compression. IIRC richard talked about circa 25% loss of efficiency @ 250% compression. IMO i would want to stay away from that area of the curve and intuitively would thus stick to 30~70% overfill
ALSO NOTE that your choice of bag design (assuming its a bag and not quilt) influences the compression of the layers INSIDE when you use them to extend your temp range-

EG - when i ordered a new bag from Tom@nunatak I essentially overfilled by approx 40~50% (its hard to say since the area is hard to calc - but i lowered baffle height by 30%, increased total fill by 5% and lowered average girth by another 5~10%, and chose 950FP down vs standard 875)
BUT by choosing a very narrow girth I not only got an warm/light bag but also integrated the fact that i will compress my MB UL down pants and jkt when I wear them inside so there is some more compression there..
Mike

Edited by mike_hefetz on 10/01/2012 16:55:12 MDT.

But, if you're talking about overfilling and maintaining the same loft, then the weight of the fabric will be the same.

If you overfill by 100%, you'll have twice as much down. The loft will be the same so the face fabric on both sides will be the same, and the baffles will be the same.

I'm just looking at the one chart "the effect of down density, in a fixed 3.25" baffle size, on the insulation value of sleeping bags"

Ugh, I wish that thread would die...or people would read all the way through it.

Richard is right...clo/inch increases up to about a 2.5x compression rate. The problem is as backpackers concerned about WEIGHT, not bulk, this is worthless to us. If you recalculate, clo/inch increases but clo/oz decreases. (The reason is weight increases faster than clo when you overstuff a baffle).

To be a bit clearer, if you double the down in a one inch baffle, the weight of the down in the baffle doubles, but the warmth of that baffle does not double (it does get warmer, just not a full 100% increase in warmth). So overfill is thermally less efficient in a static test.

So fully lofted will always be warmer per weight than compressed...in an ideal situation.

That said overstuffing does have some real world benefit; namely to ensure that down stays on top of your body instead of shifting to the sides. Factor in some fill power loss due to environmental humidity and you get the reasonable fudge factor of 20- 30% overfill (aka a few ounces extra) that you see so many manufacturers offer.

Years ago I use to work for Feathered Friends in Seattle. Calculation of overfill was always by weight and not volume. You could ask for a percentage overfill or just specify ounces for a given product. I had my own sleeping bags and jackets overfilled as a norm. I still have a sewn trough jacket that was about 30% overfill and the loft on that jacket has held up for 25 years of use. Being sewn through you can feel the cold spots easily but put a shell over it and its uber warm.

I believe that currently most quality sleeping bags are overfilled compared to what they use to be (25 years ago). Back then if you had a 2" high baffle and you calculated your down to fill that level of space if was considered properly filled because you let down reach it's maximum lofting efficiency unhindered. However, when you put a bag filled the old way next to one with even a 15% overfill of slightly lower quality down say 750 vs 800 and the 750 looked nice and plump pushing past the 2" vertical baffle, put them side by side and guess which one was going to walk out of the store first. I remember this exact conversation with Peter Hickner many moons ago.

Sorry Dustin : )

I'm pretty sure the question was if you overfill down 100%, will you get the same warmth for the weight. That is the only thing that's important so maybe I assumed that.

i.e. will the clo/oz/yd2 be the same if you compress up to 2.5x?

it's a little important because if you don't make the baffles big enough, you'll lose some warmth (clo). For example, you'll need about 30% more weight of down to achieve the same clo if you compress the down 2X.

this is offset a little because the baffle width is less - so maybe it's 25% more weight

I agree with you that "clo/inch increases up to about a 2.5x compression rate" but that's not what's important - it's the weight of the garment.

I'm pretty sure I did answer that question. I specifically mentioned that while clo/inch does increase, it does not increase at the same rate that the weight increases. Put more accurately, 1 inch of 100% overfilled down does not provide the same warmth as 2 inches of uncompressed down. Therefore both fill weights will be equivalent, but the fully lofted down will be warmer. The point is that Richard Nisely was technically correct in his analysis, he was just analyzing the wrong performance metric for the UL crowd.

So as explicitly as possible, compressing down decreases the clo/oz (regardless of any clo/inch increase).

To maximize warmth/weight, keep your down as fully lofted while minimizing down shift and cold spots...again usually a 30% or few ounce overfill accomplishes that goal.

@dustin&@Jerry
that was my whole point - richard posted MEASUREMENT RESULTS for that exact same scenario of same shell but diff fill weight
i might be missing something...but if you read the thread you will see richard post:

".... the sleeping bag's clo/kg calculation uses the total sleeping bag weight, not just the down weight.

Bag1: Down kg = .454, bag weight kg = 1.115, clo/kg = 4.0
Bag2: Down kg = .539, bag weight kg = 1.178, clo/kg = 4.0
Bag3: Down kg = .624, bag weight kg = 1.260, clo/kg = 4.0
Bag4: Down kg = .709, bag weight kg = 1.347, clo/kg = 4.0
Bag5: Down kg = .794, bag weight kg = 1.440, clo/kg = 4.0

"

I agree with you totally Dustin. You explained it very clearly.

The question isn't whether anyone is right or wrong, but how much down and loft to use to maximize warmth and minimize weight (and avoid having the down shift so you have empty baffles that result in cold spots). Or how much down do you need to be warm down to a particular minimum temperature.

I don't understand the 5 bags example, Michael. Not enough information to figure out. I'm not really into rating finished bags, I just want to figure out how much down or synthetic to stay warm down to a particular minimum temperature, and there isn't enough information out there to figure this out which is why I started measuring it myself.

Since the goal of the OP is to identify the warmest scenario, I'll add a little mud to the water...

When you overfill you risk having a bag that is "stiff" and does not "drape" well across your body. This is especially true on a bag that is generous in girth. Think of lying under a sheet of newspaper - gaps all over the place.

An overfilled form-fitting differential mummy cut might be forced into all your nooks and crannies, but you risking compression cold spots if it is to tight.

So, while some overfill may be a good thing, an excessive amount that looks good in the charts will leave you will hollows and/or cold spots in the field.

Let common sense prevail.

Wow, thanks for all the contributions. I will attempt to guide this tread to some sort of final conclusion(s). I hereby volunteer to be put in front of the firing squad:

At first glance, the value of the ‘overfill debate’ might seem a bit exaggerated: the added value of a smaller – read: lighter - shell vs. total down weight is more or less neutralized by the lower clo/ounce value of down in an overfill context. For example – and now I am cutting some corners for the purpose of simplicity - a down bag with a shell weight of 200g and a 100% overfill, resulting in 400g down in total, will more or less have the same insulation value of a bag with a shell weight of 300g – read: bigger baffles – with a non-overfill of 300g down in total.
Again, I am cutting some corners – the values in my example are obviously arbitrarily chosen without any scientific backup.
As there seems no difference in total weight vs. insulation, one might conclude that a non-overfill scenario is better for secondary reasons (e.g. quicker drying).
In this analysis, again for the sake of simplicy, I make abstraction of a small amounts of overfill, aimed at compensation humidity and filling up the last cubic inch of my baffles.

This is however a conclusion drawn to hastily, at least from the perspective of a lightweight backpacker. Let’s go back to Richard’s graph:



It shows that overfill will – up to a certain amount – result in a LINEAR increase of clo (and not a degressive increase). Assuming that a normal fill corresponds exactly with the volume of the baffles and making abstraction of external variables such as humidity, there will thus be no difference in clo/ounce down between e.g. a 30% overfill or a 80% overfill.
Or to put it differently (and more sharper): even an overfill of 10% will result in a not-negligible decrease in clo/ounce down (again, assuming that a normal fill corresponds exactly with the volume of the baffles and making abstraction of external variables such as humidity. In real world situations, 10% overfill might still result not result in a ‘clo/ounce penalty’ for the reasons already mentioned earlier in this thread).

If my analysis is correct, then my conclusion, from the perspective of someone how wants maximum weight efficiency, would be:
- The question a lightweight backpacker should not primarily be: with HOW MUCH should I overfill, but rather: should I overfill at all or not at all.
- And if you decide to go for an overfill, rather choose the SMALLEST shell available and overfill it AS MUCH AS possible, up to the amount corresponding with the desired minimum temperature rating. According to Richard’s own measurements, one can overfill up to at least 2,5x without breaking the linear increase in clo/ounce down.

Or to put it bluntly: from a lightweight backpacker’s perspective: it is an all or nothing question. Again, in my (proposition of a) conclusion, I am of course assuming that the baffle volume corresponds exactly with a normal fill and I am making abstraction of external variables (e.g. humidity), just for the sake of maintaining clarity with regard to the main issue of this thread. Bear that in mind before pulling the trigger.

Edited by wim_depondt on 10/02/2012 10:41:30 MDT.

I agree, there's a "firing squad" aspect to this : )

If you want to talk about the weight of the fabric, it of course depends, but:

Let's say you have 3 inch loft. That would be 5 ounces of down per square yard.

Let's say you use 0.9 oz/yd2 fabric for shell and baffles. Let's say you have 3 inch wide baffles. Then there would be 12 baffles 36 inches long per square yard. Add 1/2 inch on each side for seam allowance so the strips are 4 inches wide. That would be 1728 square inches of baffle per square yard = 1.33 square yards. Plus a square yard for inside and outside = 3.33 square yards = 3 ounces for the fabric. Total of 8 ounces per square yard of area. Sleeping bag = 3 square yards so total bag = 24 ounces. Plus an ounce for a zipper = 25 ounces.

If you over-filled 100%, you'de need 30% more down by weight to get the same warmth (clo), so you'de need 6.5 ounces of down per square yard and the loft would be 1.95 inches. The area of the baffles would be 1274 square inches = 0.98 square yards. Total fabric = 2.98 square yards = 2.68 ounces. 9.18 ounces per square yard total fabric+down. 27.54 total ounces for bag. Add an ounce for the zipper = 28.54 ounces. 3.54 ounces heavier than fully lofted case.

Except, I have a theory that if the loft is less, the baffle width should be less or it's harder to keep the down from shifting to one end, so in the second case the baffle width would be 1.95 inches so the area of baffles = 1.51 square yards so weight of fabric is 3.16 oz/yd2. Total weight for bag = 28.95 ounces + 1 ounce for zipper = 29.95 ounces for the bag which is 5.95 ounces heavier.

But it'll be different depending on your fabric, baffle width, loft,..., but that's an apples to apples comparison.

I think the over-all conclusion is to not over-fill if you want to minimize weight for a particular warmth. Except you have to over-fill 10% if the down is rated with American standard because it's more liberal, and you want to over-fill 10 to 15% to make sure the down doesn't shift leaving places in the baffles with no down.

And the other thing is to have the baffle width right. I have this theory that if you have 3 inches of loft, you want 6 inch baffles (twice the loft - inconsistent with above example - maybe the rule should be baffle width = loft). Wider baffles = down will shift creating empty spots. Narrower baffles = heavier than optimum. But I've never heard anyone talking about this so I might be wrong.

Need to add baffle height into the equation.

baffle height = loft

If you overfill, loft will exceed baffle height.

To look at an unrealistic example if you had a baffle material height of 1 1/2" and baffle width of 8" and you overfilled with 30% (fill weight) do you still expect a loft of 1 1/2"?

Yeah, you're right, if you overfill it will push out, I just did rough calculation.

@JERRY
im not saying im right or wrong - but if you did read that specific post form richard you would see that the data i copied (as seen in the graph wim copied here)

He MEASURED 5 exact same buid bags with diff amount of down in the baffles and the MEASURED RESULT was that the TOTAL CLO / TOTAL BAG WEIGHT stayed the same.

Now i understand the calculation you are trying to do on richard's data - but what i was trying to point out is that there is some contradiction there.
at the end of the day ALL WE CARE ABOUT is total weight per total warmth...and while your motivation to extract only the effect of the down is understandable - it MIGHT not be doable

Mike