Does not recognise acceleration load curves

[Demirism]

I have manually added the time and acceleration values into a load curve, however in the table next to the graph the title for the accelerations is 'Load Factor', then when I run an analysis it ends suddenly as it does not recognise the load curve as accelerations, rather as a permanent load. How do I get the program to recognise the accelerations?

[huffte]

Hi Demirism.

The 'Load Factor' title on the accelerogram input is not the problem.

The column titles are simply a standard way of denoting the required input.

1. Make certain that you specify the correct time column, 'load factor' column, first line value, and last line value. Do not just accept the defaults when opening the accelerogram file.

2. The accelerograms should be applied at supported nodes as acceleration loads.

3. The factor to be applied to the accelerogram load depends on the nature of the input accelerogram. If it is on g's and you want to apply the full load and the units are cm, then the multiplier in Applied Loads for the accelerogram would be 981. Make sure you don't apply this twice. If the accelerogram is already in cm/s/s, then the multiplier in the applied loads dialogue would be 1, not 981.

I suspect you are having trouble applying the permanent load, which has to be done before the response history from accelerogram loading may commence. You may want to do a forum search on 'unable to apply' or whatever the error message you get is.

Best of luck Demirism.

[Demirism]

Hi huffte

Thanks for the explanations, however:

- I created a load curve and manually added the time and acceleration values in.

- The accelerations are applied to the structural node at the bottom of my bridges columns and I've made sure the type of load when applying it to the nodes is specified as acceleration.

- I believe the accelerations are in the correct units so the multiplier I use is 1.

For my analysis, I am only using the accelerations and not any permanent loads.

Having made sure I compared my model input data to our help, I am still confused as to what I am doing wrong and why SS believes the accelerations are permanent loads?

[huffte]

Can you send the model to me via the email address listed in my forum profile? Let's see if we can determine what is going on.

[huffte]

Even with no explicit Permanent Loads specified, the masses are, by default, converted to loads. So, I am still not convinced that SeismoStruct is treating your acceleration loads as Permanent Loads, but I believe there must be some instability, either numerically or structurally, under simple gravity loading.

[huffte]

Hi Demirism. I have taken a look at your model.

You have defined inelastic 'rss' sections composed of concrete with no reinforcing for all members in the model. It is no wonder that your model cannot sustain even it's own self-weight - that is why you get the message that the permanent load cannot be applied.

You need to convert your member sections to reinforced concrete types and specify the reinforcing if you wish to maintain a nonlinear analysis. Of course you could change the element types to elastic frame elements and get it to run, but this would tell you nothing about the structure's actual response.

Also, your input accelerograms are unusual. The starting point is 0.5 seconds and the time step is 0.1 seconds. I am uncertain as to what type of load this represents, but make sure the applied load is representative of the load you are attempting to model.

Best of luck Demirism.

[Demirism]

Thanks for that huffte

I've changed the sections to rcrs and added some reinforcement however it is still coming up with the same error.

Was there anything else wrong with my model?

Thanks

[huffte]

Hi Demirism.

Yes, you have changed to reinforced concrete sections, but there are still problems.

You have put such a small amount of reinforcement in the members (roughly one-tenth of the minimum allowed by typical codes) that the members may as well still be unreinforced. This is why it won't apply the permanent load - it collapses before the self-weight of the structure is applied.

If you will increase the reinforcement patterns, the model runs. But it still collapses before the end of the acceleration history. Are the member sizes representative of an actual structure for the specified span lengths? Is the acceleration history representative of an actual, realistic ground motion? It has a maximum spectral acceleration of about 2.4 g.

I think your fundamental modeling is close, Demirism. However, the reinforcing, possibly the member sizes, and possibly the input loading are seemingly unrealistic, causing structural collapse either before the self-weight has been applied (with your reinforcing scheme) or during the acceleration loading.

Best of luck Demirism.