Rockwell Hardness as a predictor of toughness/brittleness?

Okay, this is yet another SAE Grade 5 vs Grade 8 fastener question - are Grade 8s more brittle than Grade 5 and does it matter for practical purposes.
My main question in this post is "I thought that generally harder materials were more brittle rather than tough?"
I've been going around in circles trying to definitively answer this and just get conflicting answers. Yes it is. No it is not. Yes it is but it doesn't matter because the Grade 5 will have already failed. It is also my understanding that SAE does not specify notch toughness or ductile/brittleness for Grade 5/8.
About the best explaination I've seen is at http://www.rockcrawler.com/techreports/fasteners/index.asp which is an automotive enthusiast site.
However, I disagree with what the author wrote in the second paragraph quoted below.
I'm fully prepared to be shown wrong! :-) It's been over 20 years since I've studied the stuff. :-(
Thanks
Jay
****from roughly the middle of the weblink above ****
Again, you can see that the grade 8 will support over 1000 lbs more or a 1/2-ton more. But there’s something more important to note. The grade 5 fastener has already reached its ultimate load and FAILED BEFORE the grade 8 starts to yield or stretch. Therefore, the argument that you should not use grade 8’s because they are more brittle than grade 5’s is not a true statement in most applications.
Toughness is an important feature of a fastener. It is the opposite of brittleness and gives you an idea of how it will handle abuse without being damaged and eventually weakening the fastener or can cause fatigue to appear much earlier than normal. One way to “measure” toughness is by looking at the hardness rating of a fastener. The higher the number (Brinell, Rockwell …) the harder the material is and the tougher it is to damage. According to Marks’ Standard Handbook for Mechanical Engineers, Grade 5’s typically have a core Rockwell hardness of C25-C34 whereas a grade 8 typically has a core Rockwell hardness of C33-C39. Based on this, grade 8’s are tougher than grade 5’s.
**********************************

Cool post I'd like to get to the bottom of this one too. What does SAE say about it? They are the ones that came up with their specs.
  Quoted Text Here  
He was doing ok with toughness vs brittleness but then went and left strength out, used "tough" instead. Very common. :/
  Quoted Text Here  
My question is about the "speed" of such testing. Does grade 5's and grade 8's ever switch places on strength depending on the suddeness or slowness of the stress applied?
Alvin in AZ

  Quoted Text Here  
I agree. However it does show that those whom know some, think they know it all!
The most useful method of addressing "brittleness" is to address it's functional opposite i.e. toughness denoted as Kci, Kcii, and Kciii i.e. fracture toughness desigated in one of three basic fracture modes(roughly tensile, shear, tearing).
Unfortunately I never seen a hardness (or even tensile strength) verses Kic plot(assuming room temperature) for a given grade of steel commonly used for grade 5 or 8 fasteners.
In the absence of that data a sensitivity analysis using the approximate relationship Kic=n* SQRT (2*E*YS*e/3)as presented in R.W. Hertzberg's text "Deformation and Fracture Mechanics of Engineering materials" with SAE data for grade 5 and grade 8.
I think you'll find their nearly identical (since TS*n is nearly constant for a given steel heat treated to 20 - 50 HRC), naturally this would indicate Grade 8 has greater toughness, tensile, and yield strength that Grade 5.
Ed Vojcak

snipped-for-privacy@megsinet.net (Edward D. Vojcak) wrote in message
  Quoted Text Here  
Ed:
Identical fracture toughness probably isn't good enough for the real world.
Given that bolts are inevitably going to get torqued to their yield strength (FTY), the real question should be whether a Grade 8 fastener has enough fracture toughness (KIC) to tolerate the same flaw size (A) as a Grade 5 fastener, using KIC = FTY* sqrt (A). Since the minimum yield strength for Grade 5 is 92 ksi, and the minimum for Grade 8 is 130 ksi, you would need 1.41 times the fracture toughness to have Grade 8 be better than Grade 5. I doubt that you can get it from these alloy steels.
Another problem with specifying Grade 8 bolts is long term durability in outdoor environments where corrosion may occur. The upper hardness limit for Grade 8 and similar structural bolts like ASTM A490 is literally where things begin to fall apart via stress corrosion cracking. For a fairly recent discussion see Hendrix http://www.hghouston.com/n062297.html
Pittsburgh Pete -------------------------------------------------------------- LEGAL DISCLAIMER
We don't believe what we write, and neither should you. Information furnished to you is for topical (external) use only. This information may not be worth any more than either a groundhog turd, or what you paid for it (nothing). The author may not even have been either sane or sober when he wrote it down. Don't worry, be happy.

  Quoted Text Here  
I would agree with this, I dont generally have problems with brittle fracture, as temperatures are rarely very low, however I would be concerned about high strength fasteners in very cold conditions. I am more used to metric grades. Grade 8.8 I find generally robust. Grade 12.9 and 14.9 potentially suffer from hydrogen cracking from corrosion. Given the choice I would not use grades above 8.8 in corrosive conditions.
  Quoted Text Here  

  Quoted Text Here  
For those of you confused Americans, the metric bolts in ISO standards that David Deuchar is discussing use what are called property classes. The numbers to the left of the decimal point are approximately 1/100th of the minimum tensile strength in MPa. The number including and to the right of the decimal point is the ratio of minimum yield strength to minimum tensile strength.
In the U.S. standards system the metric bolts are in SAE J1199 and the English bolts are in J429. The conversion guidance section notes that property class 8.8 is approximately equivalent to SAE (J429) Grade 5, and class 10.9 is approximately equivalent to (J429) Grade 8. In J1199 there currently is no property class 12.8 or 12.9 that are higher strength cap screws.
In the 1978 version of J1199 it said that "Caution is advised when considering the use of class 12.9 bolts and screws...Some environments may cause stress corrosion cracking of non-plated as well as electroplated products". In the 1983 revision classes 12.8 and 12.9 disappeared. In the early eighties there had been a large recall of class 12.8 bolts used on vehicle suspensions. See the reference to Hughel in: http://www.hghouston.com/herebut.html
Also see a discussion of a hydrogen embrittlement failure of socket set screws at: http://www.meic.com/Newsletters/2003/Winter03.htm
Pittsburgh Pete

Cool post, thanks Pete! :)
If you got more or get more bring it on. ;)
Alvin in AZ
  Quoted Text Here  

  Quoted Text Here  
Using the equation Kic=FTY * sqrt (A) with FTY = yield strength is "circular reasoning" since it directly implies Kic (grade 8)/Kic (grade 5) = 1.41!
The correct approach is to use the the Kic/FTY ratio (as well covered in Barsom/Rolfe "Fracture Fatigue Control in Stuctures" 2ed pg 200 - Barsom and Rolfe use "assumed Kic" values for the analysis. As I said Kic vs hardness, Tensile (or yield) strengths are not available: If your going to do any assuming why not just assume: tensile strength Kic/SQRT(A) where the fault size is very small (on the order of 5 µm)and fracture toughness is not an issue. Hence Grade 8 is "stronger" than grade 5.
Ed Vojcak

snipped-for-privacy@megsinet.net (Edward D. Vojcak) wrote in message
  Quoted Text Here  
What Ed started hand waving about was an equation shown on page 396 of the 3rd edition of R. W. Hertzberg's "Deformation and Fracture Mechanics of Engineering Materials.
It is K1c = n SQRT(2*E*YS*e/3)
where n is apparently the coefficient in the power law expression relating true stress, sigma, to true strain, which is sigma = K (strain)^n (on page 18) E is the elastic (Young's) modulus, YS is the yield strength, and e is the true fracture strain in uniaxial tension.
For a round tensile specimen e = ln(1/[1-q]) where q is the reduction in area expressed as a ratio rather than a percentage.
I interpreted what he said as being that he thought that grade 8 somehow had greater toughness than Grade 5.
I have not been able to find any KIx (not plane strain) data for SAE Grade 5 and Grade 8 bolt materials. The closest thing seems to be some ~30 yr old work on structural fasteners by Townsend shown in Figure 4.28 of the "Guide to Design Criteria for Bolted and Riveted Joints", 2nd edition which can be downloaded at: http://www.boltcouncil.org/guide1.htm That data shows KI increasing as hardness decreases from 48 to 35 HRC. I don't see why it wouldn't continue to increase as hardness decreases further to Grade 5.
Picking the same critical flaw size is one possible choice for comparing fasteners. It just assumes that the manufacturer is too lazy to change his ndt setup when he changes grades.
Pittsburgh Pete

  Quoted Text Here  
Well Pete,
I was not aware that is was wavying any hands - but addressing a question most progmatically. I agree that there is little data on Kic verses YS and thats why extrapolation is wrong. Best optimized engineering analysis is done using the Kic/YS ratio. Given the correct high level of stress Grade 8 fasteners would be more fatigue resistant than grade 5 and may never develope a critcal flaw through a fatigue mechanism as a grrade 5 fastener would.
BTW it not a matter of someone being "too lazy to change his ndt (sic NonDestructive Testing) setup" but one of POD (probability of Detection) maybe you can research that issue with in ASNT publications. We all know how good your reseach is.
Ed Vojcak