Short ED and the substantiation thereof.

[koopas]

Hello everyone,

I have a question regarding short ED conditions and their substantiation. The following is the 3-step method generally accepted:

1. First, one either looks up or computes the bearing strength of the fastener/material combination using the SRM tables or the bearing formula Fbru * d * t, respectively. Note the fact that the SRM tables are based on 2D ED and that the Fbru used in the bearing formula in this step is Fbru at 2D ED (some people use the Fbru at 1.7D ED).

2. Then, one computes the shear-out allowable based on the short ED condition (for instance, assume our discrepant fastener only has 1.5D ED). Shearout formula used is Fsu * 2 * t (ED - 0.383D) where Fsu is the fastener's.

3. If the value in (2) is greater than (1), the short ED of 1.5D is acceptable for strength.

My question is this:

The calculation in (1) seems to be in error since Fbru in the formula "Fbru * d * t" is a function of edge distance. Likewise, the SRM tables are only valid for 2D ED; with a shorter ED yielding a smaller allowable. However, in step (1), we looked up (or calculated) the bearing allowable based on the ORIGINAL joint configuration with an edge distance of 2D, when in reality our actual discrepant fastener possesses only 1.5D ED. This shorter ED of 1.5D will incidentally yield a smaller bearing allowable than the one looked up or computed in step (1).

This renders the comparison in step (3) invalid and illogical, since you're now comparing a shear-out allowable (albeit based on the shorter ED of 1.5D) to an allowable that's now lower than the original design joint bearing strength as explained in the previous paragraph. Since a joint is inherently designed to be bearing-critical, you've just demonstrated that the short ED condition yields a bearing allowable that's now lower than originally designed, regardless of the new shear-out value calculated in (2). This cannot be acceptable for strength!

Could you please shed some light on this? If the above procedure is not correct, could you please describe your method of substantiating short ED conditions?

Thanks,

A confused Alex

 

[jetmaker]

koopas,

Actually, in step 1, you are suppose to use the Fbru for the e/D that the part actually has. So, if you have an e/D of 1.5, then look up the fastener capabilty value in the table, see whether that value is based on an e/D of 1.7 or 2.0, and then ratio the material Fbru at e/d=1.5 to that used to compute the table (i.e Fbru e/d=1.7 or 2.0).

This is the value you compare with step 2. Use the lower of the 2.

Then calculate the capability lost by comparing the lowerest value calculated against the per drawing fastener capabilty. You can sometimes wiggle the per drawing capability by using minimum drawing tolerance conditions.

Hope this helps.

jetmaker

 

[koopas]

G'day Jetmaker,

The procedure I initially listed is the one followed in a Douglas training manual for sustantiating short ED conditions.

In step (1), the bearing allowable that's either looked up or calculated, is based on a FULL 2D ED, even though the discrepant fastener only has 1.5D ED.

In step (2), the shear-out allowable is based on the shorter 1.5D ED, and the subsequent step compares the latter to the bearing allowable found in step (1). If the shear-out value in (2) is greater than the [questionable] bearing allowable in (1), the short ED condition is acceptable AS-IS for strength.

From what you're writing, the new bearing allowable found in step (1) will incidentally be lower than the 2D ED bearing allowable by virtue of the fact that the ED has gone from 2D to 1.5D. The shear-out allowable has also decreased due to the shorter ED. I agree with you.

Since the 1.5D ED joint is now weaker in both bearing and shear-out when compared to the 2D ED, non-discrepant, condition, it seems impossible to justify the short ED condition as acceptable for strength, as-is, without oversizing adjacent fasteners to pick up the lost bearing load.

I am quite confused, as it seems that the widely used short ED substantiation, as published in the Douglas training manual, is meaninglesss since it does not take into account the shorter ED in step (1) when calculating the bearing allowable. In fact, as I've mentioned, it demonstrates that the 1.5D ED condition yields a joint that's weaker in all respects (bearing & shear-out).

In conclusion, step (3) of the method uses the allowables found in steps (1) and (2) and compares them against one another, when in reality these two allowables are LOWER than the 2D design bearing allowable. Isn't this comparison worthless?

I think I am missing something....your feeback is welcome.

Have a good weekend,
Alex

 

[jetmaker]

koopas,

I find it hard to believe that the training manual would make such a gross mistake, but stranger things have been known to happen.

As I have always understood it, the capability of an actual e/d of 1.5 fastener will be worse than an e/d of 2.0, except if the fastener is shear critical.

So unless you know/calculate the loads, oversizing will be necessary.

However, I will add these opinions. 1) It is possible, that Douglas does not consider a decreasing e/d to have an influence on bearing strength (not likely, but again stranger things exist); 2) that shear tear-out is generally more critical than bearing failure for geometrically proportionate parts when e/d is < 2.0 (I have no physical substanciation for this statement).

jetmaker

 

[koopas]

Hola Jetmaker,

I agree with you that the Douglas analysis must assume that a shorter ED does not reduce the bearing strength. Otherwise, the analysis makes no sense.

I've emailed the instructor, and have communicated to him my concerns.

Could you please re-explain what you meant in your second point? What does "geometrically proportionate" mean?

IMO, the loss of bearing strength due to the shorter ED is unacceptable AS-IS.

Anybody else with their $0.02?

Alex

 

[jetmaker]

koopas,

By "geometrically proportionate" I imply that the D/t, e/D, fastener pitch and gage, etc... are all within the manufacturer's design criteria.

Douglas may have set their design parameters based on Fsu and Fbru allowables such that the joint is always a shear-tearout as opposed to a bearing failure. This would allow them to simplify their analysis time by eliminating one check. Like I said tho, there is not basis for this statement other than assumption.

Regards,

jetmaker

 

[HSThompson]

Koopas,

You asked for ¢2,... well here it is....

First of all, I think you probably are correct in that Douglas has some sort of error in their formula. The stress group of the current Douglas owners (or ownees), use an Fbru based on 1.7e/D, >ON ORIGINAL NEW DESIGN<. This is/was to account for manufacturing error above and beyond what 2D+.05 provides.

Being aftermarket support as you are you have no access to the analysis reports, so margins are unknown and consdered 0.0+. Therefore you have to follow the procedure you gave with Fbru based on the edge margin you have.

If after calculating fsr shear, bearing, tear-out, and net tension the failure mode remains unchanged then your good. If not, then you have to make it up somewhere.

When you lack known margins, equivalent or better is your only option.

 

[koopas]

HSThompson,

Thanks for your feedback.

You wrote:

"If after calculating fsr shear, bearing, tear-out, and net tension the failure mode remains unchanged then your good. If not, then you have to make it up somewhere."

Since a fastener joint is usually designed to be critical in bearing, you are bound to lose bearing strength with a shorter ED.

As such, the substantiation of using the weakened joint, AS-IS, is impossible.

What do you think?

Alex