Alpha cr values?

[RRufino]

Dear all,

I'm trying to understand this new feature on ARSA 2015, alpha cr values calculation.

I've made a simple portal frame which is from bibliography, see attachement. I was trying to calculate alpha cr x but i get N/A.

Can you explain me what i'me doing wrong?

Thanks in advance,

RRufino

[Pawel.Pulak]

---

@RRufino wrote:

I'm trying to understand this new feature on ARSA 2015, alpha cr values calculation.

I've made a simple portal frame which is from bibliography, see attachement. I was trying to calculate alpha cr x but i get N/A.

Can you explain me what i'me doing wrong? 

---

Hello,

It is the effect of some limits when calculating other results for stories.

alpha cr X and alpha cr Y are calculated from other results for stories using the following formulas for each story:

(FX/FZ)* (Height/ dr UX) and (FY/FZ)* (Height/ dr UY) respectively

All these results or data are displayed by default in the Stories table or they can be added to "Stories", "Displacements" and "Reduced forces" tabs of this table.

As you can see for your model dr UX and dr UY displacements are N/A (not available) - that is why alpha cr are N/A too

And dr UX and dr UY displacements are N/A for your model because there are no horizontal elements (beams or slabs) on the top level of the story.

It was reported to development team to modify it and to calculate such displacements also in case of inclined rafters, when only nodes exist on the level of the top of the story.

I have noticed that there are also problems with the availability of results for stories for load combinations reported as "Combination of buckling cases" (buckling analysis defined for load combination). It will by analysed by dev team too.

The workarounds:

• for the first effect: manual calculation of alpha cr from above formulas or adding any horizontal element (even the dummy one) at the top node of the story
• for the second effect: for the case 5 using "Change analysis type" and selecting Statics

---------------------------------------------
If this post answers your question please click the "Accept as Solution" button. It will help everyone to find answer more quickly! 

Regards,

[RRufino]

"for the first effect: manual calculation of alpha cr from above formulas or adding any horizontal element (even the dummy one) at the top node of the story"

Thanks pp2008 but i might not understood well because i've tested it without sucess, can you make it and resend?

2 more questions:

- For a single portal frame alpha cr x shoul be equal to critical coeff. with Buckling analysis, right?

- For "Frame 2D Design Layout" i'm getting expected values to critical coeff. but changing to some other 3D layout i'm not getting it because out of plane modes...how should i workaround this (i already increased number of modes and even place some virtual supports on nodes).

Thanks,

RR

[Pawel.Pulak · Accepted solution]

---

@RRufino wrote:

"for the first effect: manual calculation of alpha cr from above formulas or adding any horizontal element (even the dummy one) at the top node of the story"

Thanks pp2008 but i might not understood well because i've tested it without sucess, can you make it and resend?

---

I have attached your model with dummy horizontal bars added. See the effect of this change on the screen capture below:

---

@RRufino wrote:

2 more questions:

- For a single portal frame alpha cr x shoul be equal to critical coeff. with Buckling analysis, right?

- For "Frame 2D Design Layout" i'm getting expected values to critical coeff. but changing to some other 3D layout i'm not getting it because out of plane modes...how should i workaround this (i already increased number of modes and even place some virtual supports on nodes). 

---

Attached the compressed file containing 3 modications of your model:

1. 2D
2. 3D
3. and 3D with support constraints corresponding to 2D.

Comparing critical coefficients of buckling analysis it can be seen that first 6 buckling modes of models 1. and 3. correspond to buckling modes 8, 9, 12, 13, 14 and 15 in model 2.

---------------------------------------------
If this post answers your question please click the "Accept as Solution" button. It will help everyone to find answer more quickly!

Best regards,

[Tuctas]

  Be aware that the calculation of acr is subjected to some limitations due to its simplicity. Read what EC3-1 says in 5.2.1(4). So, generally don't expect acr to be absolutelly equal to the critical buckling coeficient...

  However, it should be calculated also in such cases of not horizontal beams at the upper level (and also the designer to keep in mind of these limitations, maybe the relation 5.3 of EC3-1 should also be automatically calculated in Robot...). 

[RRufino]

"Comparing critical coefficients of buckling analysis it can be seen that first 6 buckling modes of models 1. and 3. correspond to buckling modes 8, 9, 12, 13, 14 and 15 in model 2."

Thanks pp, but perhaps i'm getting confused with theorical concepts...i was expecting that alpha cr x (with all workaround and putting case 5 as static) will lead equal to  minor critical coeff. (i'm wrong right?)

After writing last post i've made exactly a test with constraints and i got the same results 😉

Now, imagine a 3d model with several frames (typical industrial building)...how can i figure which buckling modes i should take in account to a certain frame? (seeing buckling modes "displacement" seems me some how a fallible method). If we regard that sometimes, due to equipment, vertical load on roofs is not equal on all roof it leads to different behaviour for different frames, the same if spacement between frames is not the same.

Thanks,

RR

[RRufino]

PP2008 can you take a look to my last post?

Thanks,

RR

[Pawel.Pulak]

---

@pp2008 wrote:

---

(...) 

The workarounds:

• for the first effect: manual calculation of alpha cr from above formulas or adding any horizontal element (even the dummy one) at the top node of the story
• for the second effect: for the case 5 using "Change analysis type" and selecting Statics 

---

No workarounds necesary for it in 2015 release - both effects were fixed in Service Pack 3 for RSAPRO 2015.

Regards,