> >
>

What exactly constitutes a bouncy floor?

#1
03-27-05, 02:03 PM
Member
Join Date: Mar 2005
Location: Chesapeake, VA
Posts: 340
What exactly constitutes a bouncy floor?

What exactly constitutes a bouncy floor? 1/4" deflection when you walk? 1/8"?

I'm designing my addition and have mined the span tables from several engineered products, calculated my loads, and the deflections that I'm coming up with just seem... well... too large. When a span is built to the maximum a span table allows (L/240, 360 or 480), it generally works out to about 1/2" total. And a walking man will create a significant % of that if the span tables are accurate. Are the floors beneath our feet actually deflecting that much and we just don't notice? Or do other things (3/4" T&G and 1/2" gypsum) provide reinforcement that isn't taken into account in the span table?

My addition has spans of 17' and 21' and I plan on using engineered I-Joists because of the distance; NASCOR NJH-14s @16"OC to be exact. My spans are right at about the max for these joists and I've run calculations- I can live with a static deflection of 1/2" over a 21' span. A standing man will defect it 1/8" and a walking man probably twice that, though... so, will that be a noticibly bouncy floor? The span tables say "spans are based on composite action of 3/4" thick sheathing nailed and glued to joists". The EI and K values I used to calculate deflections aren't, though. Does that mean that this is a true deflection? Or does that still assume a single otherwise unsupported I-joist, and the true reaction of a man walking will be spread by the 3/4" ply onto 2 or 3 joists and lessen the reaction? I should note that plugging a (max) 21'-10 span into the straight engineering calcs yields a deflection fully 3x the L/480 the span tables are supposedly calculated for.

Thanks!

#2
03-27-05, 03:02 PM
Member
Join Date: Feb 2002
Location: United States
Posts: 4,679
grover,

Don't know if you are trying to be an engineer or are one but lets make it simple. For the average consumer, the task that you are undertaking in calculating this all out is based on theories, tested under certain conditions. I always go one size up from what I need to accomplish. All at L/360. It may seem large to you but when you have been in the business, you soon find out that what you think might work, won't.

Draw out what you are designing and let a professional determine what it is that will accommodate the span. Your local lumber yard has company reps tht can determine this accurately and will do this for you, in most cases. This is what will be done anyway. This way the project is designed properly and the engineers stamp provides the assurance that all is right for that application.

Its the best advice I can give you. Hope this helps!

#3
03-27-05, 03:24 PM
Member
Join Date: Mar 2005
Location: Chesapeake, VA
Posts: 340
Thank, Doug. I am an engineer, actually, but don't really have any practical experience with regard to joist engineering. I got the span tables & engineering data for the I-joists from the local distributor, and discussed the spans with them; they recommended, as you did, to oversize. I'm quoting numbers from the 40/25 simple span table, even though my actual values will be closer to 30/10 and the spans for the bedroom side of the addition will be continuous over a central bearing wall, which *should* give a little leeway, even though I'm within a foot of the max span. (Is that enough? Or should I oversize more?) The family-room side of the addition will be using 14" I-joists (rated for 21'-10" span) for a 17' span, so they should be 2.5x stiffer right there. Incidentally, I DID bump my floor from 12" to 14" I-joists- though it was moreso because they didn't stock the wide-flange 12" I'd need, and the medium-flange wouldn't make it. The 14" is considerably stiffer.

I didn't even get into the deflection calculations until my wife decided she wanted a giant elliptical opening around the staircase which became difficult to support, and I had to really run some numbers to be able to support a 3600lb load in the middle of a 16' span. Luckily the engineering data was easy to come by and I've done everything with very conservative calculations; I'm just concerned about translating numbers on paper into real-world subjective performance... That is to say, I'm very comfortable about it structurally, I just don't want to build it only to have it be uncomfortably bouncy.

I guess what I'm getting at is that it really shouldn't matter what's *in* the floor, being 2x12s or 2x8s or 16" LVL; how much deflection is noticible when you walk on it?

Interestingly enough, the span table had the same maximum span value at 21'10" for this 14" I-joist loaded at both 40/10 and 40/25...

#4
03-27-05, 03:34 PM
Member
Join Date: Feb 2002
Location: United States
Posts: 4,679
grover,

I'd have to look at to see what you are trying to do. Unsure as to exactly what you are wanting to do. I assume you are trying to avoid a support post. I have done arched (open to above) projects. It would be my assumption that possible going to a web floor truss would be an alternative but unsure of your application.

As I and they said, oversize is the best way but without knowing more at this point, with what you just wrote I am not comfortable with providing much of an answer to your and my approval.

#5
03-27-05, 04:56 PM
Member
Join Date: Mar 2005
Location: Chesapeake, VA
Posts: 340
Thanks Doug, I appreciate the help! I thought it was going to be a simple answer of "design to 1/8" deflection" or something along those lines, but I guess nothing is simple when you start talking about anthropomorphic engineering, huh? The roof will be scissors truss, so I shouldn't need joists parallel to the roof trusses to carry and tension, right? Because of the wall through the center, the long axis actually becomes the shorter span, and by using I-joists, it eliminates the need for a beam and posts in the middle of my garage, which I like. If you're willing to take a loot at it, here's a plan view of the 1st floor, with the I-joists overlaid @16OC in green (all are Nascor NJH-14 except the two labeled as LVLs):
1st floor w/joists overlaid

The I-joist span table lists 21'-10" as the max simple span for both 40/10 and 40/25. I've added extra I-joists to support the dead load of the two (41plf) walls that run parallel to the joists. (er, actually, looks like I only copied the one other onto this view, but I intend to support it like that.)

3D View of 2nd Floor

The longest spans are over the garage, and support the bedroom/bathroom/hobby room/storage area. The area I'm more worried for bounciness about is the area over the kitchen, though, since we'll be in the living room all the time and I want it to be very stable. It will all be carpeted on the 2nd floor, if that makes a difference.

Last edited by grover; 03-27-05 at 05:10 PM.
#6
03-27-05, 05:07 PM
Member
Join Date: Feb 2002
Location: United States
Posts: 4,679
grover,

Got them. Let me review. You get my PM?

#7
03-27-05, 05:14 PM
Member
Join Date: Feb 2002
Location: United States
Posts: 4,679
Grover,

Can you save this in a DXF or DWG format? - the black background is not an easy view.

#8
03-27-05, 05:23 PM
Member
Join Date: Feb 2002
Location: United States
Posts: 4,679
Grover,

What is the reason for 14" TJI's?

What are you using for HVAC?

#9
03-27-05, 05:38 PM
Member
Join Date: Feb 2002
Location: United States
Posts: 4,679
Grover,

Depending on answers to the other posts

I came up with this L/360 -

Web Truss by SpaceJoist - 3 1/2" Web Heel/Top

14" High - 16" O.C. Span - 27'10"

TJI - 230 Series - 2 1/16" Web Heel/Top

14" High - 16" O.C. Span - 23'-9"

#10
03-27-05, 07:11 PM
Member
Join Date: Mar 2005
Location: Chesapeake, VA
Posts: 340
I'm only using TJIs for the span [edit: I see what you mean about the metal-web space joists being better for HVAC!] The local supplier here doesn't carry 3.5" flange TJIs, just the 2.5" ones which is a shame, as the 3.5" flange were my first choice. L/360 for a 21'-10 span is 3/4". Is that comfortable? Nascor's span tables were for L/480 ~ 1/2", so the span tables look very similar with that taken into account. The LVL presents a problem to the HVAC, as they have a max 2" hole... I would have to triple up TJIs vice a 3.5" LVL for the beam by the stairs to route a 6" duct through it; is tripling TJIs an acceptable practice?

I really need to have an HVAC professional take a look at the HVAC plans as I'm not at all comfortable with my HVAC skills, but I plan on contracting that part out and will definately be talking to the guy once I have the design further along. At present, I penned in a 3-ton unit pushing 1200cfm to 13 vents, 11 on a 6" flex duct feeding a 100cfm 3x10 radiator, and 2 50cfm. The duct through the garage would have to be all sheet steel to meet fire codes and that will just be out in the open.

I'll email you the dwg- I *really* appreciate your help & feedback! I'm in Chesapeake, VA, btw, so the temperate zone is fairly moderate but I do have to take wind loading into account for hurricanes. And I just finished reading the IRC cover to cover which, oddly enough, I enjoyed. Is that sad or what?

Last edited by grover; 03-27-05 at 08:09 PM.
#11
03-28-05, 03:45 AM
Member
Join Date: Mar 2005
Location: Chesapeake, VA
Posts: 340
I'm curious- what criteria determine the maximum span of a truss/joist system? The SpaceJoist 420-series Inverted Truss is rated the same (24'-0) for L/480 as L/360 for 40-20-5. So deflection alone can't be the only criteria.

#12
03-28-05, 07:58 PM
Member
Join Date: Mar 2005
Location: Chesapeake, VA
Posts: 340
I pestered the poor sales girl at the local UFP truss factory for a good 20 minutes until one of their engineers came out and ran some numbers and answered all of my questions. The good news is, I've got great numbers in-hand for TJIs. The bad news is, I still won't know if it's acceptable. (Oh, and for more bad news, I found out the joists are more expensive than I thought. And strong-ties apparently squeak, which threw another of my plans out the window...) I sent SpaceJoist an email asking for more info on their open-web steel trusses, too; hopefully they'll get back to me. UFP's lightweight open web trusses unfortunately aren't strong enough for my span at 14 or 16" depth, and their better trusses are really expensive. They don't carry any steel web trusses there. There's another lumber yard that I might be able to special order from, I haven't asked them specifically yet.

NJH-14 @16"OC, 40/10, 21.8'=0.54" Live deflection (L/482) and 0.79" total deflection (L/330). (Max span=21'-10 @ L/480)
NJH-16 @16"OC, 40/10, 21.8'=0.41" Live deflection (L/646) and 0.59" total deflection (L/440). (Max span=24'-1 @ L/480)

I looked into getting a steel I-beam and just running 2x10 joists, but until it's all said and done, there really isn't much (if any) cost savings over the TJIs, and I'd end up with a giant steel I-beam exposed in my garage. Especially with the price of steel going up like it has! Maybe I can just get some sky hooks off ebay?

#13
03-28-05, 08:21 PM
Member
Join Date: Feb 2002
Location: United States
Posts: 4,679
grover,

Just to let you know, I will have the price tomorrow on the steel web trusses but the price I got on the TJI's were \$2.20 per LF.

I have never had any simpson hangers squeak - that is if they are installed right. They don't make any noise at all, not on my jobs.

Another day to wait for answers

Last edited by Doug Aleshire; 03-28-05 at 08:52 PM.
#14
03-29-05, 03:35 AM
Member
Join Date: Mar 2005
Location: Chesapeake, VA
Posts: 340
What's the "right" way to install them? He said a lot of installers put construction adhesive on the hanger as well as nailing them to minimize the squeak, but that there is still metal-to-metal contact between the nail and hanger and that squeak is inevitible as the joint is stressed/unstressed. He may have been referring to the long-term, as the nails loosen over time.

I was quoted \$2.27/lf for their version of the 14" TJI, \$2.50/lf for the 16" and \$1.86 for the 12".

#15
03-29-05, 08:15 PM
Member
Join Date: Feb 2002
Location: United States
Posts: 4,679
grover,

I did get price on the SpaceJoist TE, as well as spoke to the President of the company in Texas.

In regards to your pricing of the TJI's, the Metal Web Floor truss - 14" 320 Series, it would be about \$2.20-\$2.30 LF. Surprisingly not too much different than the TJI's. I did notice however that if we go to the 420 Series which is rarely used in residential, the cost is much more.

I have seen those use construction adhesive and this is fine. However I have seen 8d or short roofing nails installed on the hangers versus the recommended 10d. With most products it is always installation errors that are the problem. The other issue is the lack of construction adhesive and using only nails to the subfloor. I insist on screws and adhesive. Once this system is interlocked, metal hangers in my opinion cannot work themselves out as you described.

Hope this helps!

Last edited by Doug Aleshire; 03-29-05 at 08:37 PM.
#16
03-30-05, 04:52 AM
Member
Join Date: Mar 2005
Location: Chesapeake, VA
Posts: 340
Well, if SpaceJoists are availible around here, they do sound like my best option Thanks!

#17
03-30-05, 05:59 PM
Member
Join Date: Aug 2003
Location: usa
Posts: 406
hey guys - you gots a complex discussion going on.

here's the rule of thumb about 'bouncy floors'

first, deflection is based on all applied loads - and the code will tell you that. I don't think you, as a person walking across the floor, will hit the maximum load, unless you are a REALLY BIG GUY!!

the code requires dead load (of course) plus live load. the total of this is quite a bit more than you weigh. in practice, you shouldn't notice any deflection in residential applications unless you've UNDERSIZED the joists.

so - what is "too much" deflection? I think it depends on span - L/360 might be only 1/4" on a "short" span, but it might be several inches on a "long" span.

my feeling is - if you notice the deflection when you walk, you won't feel that the floor is strong enough - even if it is strong enough by code.

I'd check your numbers again in terms of applied loads and deflection - 1/4" deflection is fine - but 1/4" of "bounce" when you walk is not.

#18
03-30-05, 07:58 PM
Member
Join Date: Mar 2005
Location: Chesapeake, VA
Posts: 340
Well, it's a 21' span- there's going to be deflection... In the living room area, I ran calculations by two methods. One was the strict formula for the I-joist using the EI & C from the tables and neglecting composite action- that said that a 200lb man standing on a bare TJI will deflect the joist just under 1/8". But that formula also shows the beam with full load would deflect about 3x what the L/480 column implies, so I know it's not entirely accurate. Assuming the span tables can be scaled by a factor of L^cubed, reverse-calculating the deflection for a 200lb man in my living room gives about 3/64" deflection (And 1/16" on the longer span). And of course more if I jumped up and down at center-span But really, I have no basis for comparison- I don't recall any residential 2nd floor moving under my feet, but I have no idea what the deflection in those floors were to know how 3/64" or 1/16" compares. Are they negligible, do you think?

I have a feeling resonance would be the real killer here, but that it's damned near impossible to calculate and not much I could do about it anyhow...

Unrelated question: Do TJIs or Spacejoists come pre-cambered for the dead load?

#19
03-30-05, 08:50 PM
Member
Join Date: Feb 2002
Location: United States
Posts: 4,679
Are we done yet?

I have to agree with trance - it's apparant that the question has been addressed and answered. We could keep debating in a friendly manner but this thread is getting too long. We all could go to school to learn about all the methods of calculating load, deflection, etc, and try to understand the theories but is it really necessary? Not really and not here.

What it comes down to is this. Various products are available to the consumer in construction of their projects. Beams, LVL, glulams, floor trusses, roof trusses, etc. are all pre-manufactured by good companies that will review the plans drawn and validate them or change what is drawn to meet the structural requirements needed. The major wood components are reviewed by the engineers before they are even made. They have the engineers that do this stuff daily. They will provide the stamp of approval on them to be presented to the Building Inspectors at the time of required inspections.

In practice, it is better to overbuild than underbuild. The bottom line is cost. If you are willing to pay for a product that will meet your expectations, then do it. If you try and save a buck or two, there are other products available and/or methods to build that might cost less. Each product can perform at or above ones expectations but only you can decide the best way.

Part of the designing that seems relatively easy, is laying out all the rooms (floor plan layout). Hard part comes in making choices of how and what products will be best suited to construct it. In conjunction with that, then you have to consider the costs. Design and drafting isn't just a click here or there. All the issues are thought out before that final set of plans go to the lumber yard. Sometimes seeking a structural engineer beforehand is the best avenue in getting the right answers for those technical issues.

What is the best way may not always be the cheapest way but it will be the right way.

#20
03-30-05, 09:49 PM
Member
Join Date: Aug 2003
Location: usa
Posts: 406
well, remember - you are not standing on just 1 joist - the subfloor is spreading your weight to at least the 2 neighboring joists. "tributary area"

and code requires something like 50 pounds per square foot - on every square foot of the floor - to determine deflection. obviously, you do not load the floor that much in normal use, and so your 200 pound point load, while higher at a single point, isn't very much in terms of the total load the floor can take.

remember, deflection is different than "bounce".

resonance is an issue in large buildings - where the results of movement can be measured in full inches or even full feet. in these cases, or even in the case of your house, the answer is to put a secondary structure with a different resonant period. we acomplish this in wood framing with our bridging and with the subflooring.

obviously, if you plan on having big dances on your floors, you may want to consider beefing up the floor, and making sure you don't get an artificial resonance with the beat of the music!!

don't worry too much - wood framed structures are fairly foolproof!!!

#21
03-30-05, 10:23 PM
Member
Join Date: Jan 2005
Location: United States
Posts: 455
Deflection <> SpringBoard

Hi Grover,

As Doug and Trance have said. Deflection is more a matter of long term rather than a short term phenomenom. If you feel the effect as you are walking then its short term and the calculations you have been performing are not about that at all.

Don't think that a Steel I Beam will eliminate bounce. It may actually be more springy than a wood beam. I worked for an Architect in Santa Monica back in the early 1970's and we designed Wilt Chamberlain's house. There was one location where 2 steel beams cantilevered out to a point supporting an interior sitting room above the entry. These were about 12-14" deep with a 10' cantilever at the apex. It was so noticably springy that we wound up having them removed and it was reframed with 10x14 timbers. No Spring at all but the deflection was about 1/2". This wasn't a problem - we just used a self leveling concrete fill and nobody could tell.

#22
03-31-05, 03:32 AM
Member
Join Date: Mar 2005
Location: Chesapeake, VA
Posts: 340
Thanks all, I do feel better new

#23
03-31-05, 11:24 AM
Member
Join Date: Jan 2005
Location: santa Rosa, California
Posts: 194
Grover,

you have another option in all this mass of confusion. If the deflection seems to great you can use a paralam or glu lam at the middle of the span parrellel with the joist direction you were trying to span and run the joists the opposite direction. If you decide to continue forward with the plans as they are you can always cross bridge or solid block at either the center span or 2 rows at 1/3 and 1/3 of the span. Cross bridging is the best to remove spring and bounce but both work.

good luck.

brian garrison
gerneral contractor/professional building designer.