Oggi non ci sono compleanni. |
|
2 membri (Masare2, 1 invisibile),
523
ospiti, e
1
robot. |
Chiave:
Admin,
Mod Globale,
Mod
|
|
|
Iscritto: Mar 2004
Posts: 8,282 Mi piace: 4
M Member
|
M Member
Iscritto: Mar 2004
Posts: 8,282 Mi piace: 4 |
Prego, è stato utile riprendere questo argomento generalmente ignorato. Dopo varie, forse troppe iterazioni con GPT5, tanto che adesso mi fermo, siamo arrivati a un punto credo sufficientemente plausibile. GPT5 propone 3 soluzioni, con vari livelli di conservativismo. Envelope-Fitting Bounds on Flat Rock Surface Synthesis: Simple Envelope-Fitting Bounds for a Flat Rock Surface When fitting a straight Mohr–Coulomb line to the curved Hoek–Brown envelope over 0 ≤ σ′₃ ≤ σ′₃₋ₘₐₓ , you need a practical choice for σ′₃₋ₘₐₓ that doesn’t rely on knowing the ultimate capacity or running finite-element models. Three widely used, flat-surface bounds are: Constant-Fraction of Intact Strength Formula: σ′₃₋ₘₐₓ = 0.25 · σci Rationale: Hoek & Brown (1980) observed that, for shallow loads, the minor principal stress range in jointed rock lies between 0.1·σci and 0.4·σci. A midpoint of 0.25·σci gives a representative upper bound without further data.
At-Rest Horizontal Stress (K₀) Approach Formula: σ′₃₋ₘₐₓ = K0 · γ · Df, K0 ≈ 1 − sin φ′ Rationale: Prior to loading, a rock mass at depth Df supports a vertical stress γ·Df . Taking its natural horizontal stress σh′ = K₀·γ·Df as σ′₃₋ₘₐₓ ties the envelope limit to the actual in-situ confinement.
Mid-Envelope Point via Rock Mass Parameters Formula: σ′₃₋ₘₐₓ = (s / mb) · σci
mb = mi · exp((GSI − 100) / (28 − 14D))
s = exp((GSI − 100) / (9 − 3D))
Rationale: The term (mb·σ₃′/σci + s)a shifts the Hoek–Brown curve from intact rock (s = 1) toward jointed mass (s < 1). Setting mb·σ₃′/σci = s locates the stress where intact-rock and joint-mass contributions are equal—an intuitive “mid-curve” bound.
Conservatism Comparison Method | Typical Value* | Conservatism (Lower σ′₃₋ₘₐₓ ) |
---|
0.25 · σci | 12.5 MPa | Moderate | K₀·γ·Df
(K₀≈0.5, γ=25 kN/m³, Df=2 m) | 25 kPa | High | (s / mb) · σci
(e.g. s=0.01, mb=5, σci=50 MPa) | 0.1 MPa | Very High | *Example assumes σci = 50 MPa , φ′≈30° ⇒ K₀≈0.5 , γ=25 kN/m³ , Df=2 m , GSI=50 , D=0.5 .
"Data speak for themselves" -Reverend Thomas Bayes 1702-1761 P(Ai|E)=(P(E|Ai)P(Ai))/P(E)
|
|
|
|
Iscritto: Mar 2004
Posts: 8,282 Mi piace: 4
M Member
|
M Member
Iscritto: Mar 2004
Posts: 8,282 Mi piace: 4 |
Adesso, è ora di interrogare la quarta AI, Deepseek, la opensource cinese. La sua risposta è molto interessante ed equilibrata. Hoek-Brown Criterion for Shallow Foundations Determining σ'₃max for Shallow Foundations Using the Hoek-Brown Criterion Based on the 2002 Hoek-Brown criterion and a review of geotechnical literature regarding shallow foundations, a mathematically and logically plausible value for σ'₃max in the context of shallow foundations can be derived as follows: 1. Context of σ'₃max in Shallow Foundations In the Hoek-Brown criterion, σ'₃max is the maximum confining stress up to which the equivalent Mohr-Coulomb parameters (c', φ') are calibrated to match the Hoek-Brown failure envelope. This value depends on the stress range relevant to the geotechnical problem: - For tunnels: σ'₃max is related to the in-situ stress and tunnel depth.
- For slopes: it is related to the slope height and rock mass strength.
- For shallow foundations: the relevant stress range is typically near the ground surface, where the minor principal stress σ'₃ is low, but it increases with depth below the footing.
2. Proposed Approach for Shallow Foundations For shallow foundations, the failure mechanism involves a bearing capacity failure beneath the footing, where the confining stress σ'₃ increases with depth but is generally limited by the overburden pressure and the width of the foundation. A widely accepted approach is to relate σ'₃max to the vertical stress at a depth equal to the width of the foundation (B) below the base of the footing: σ'₃max ≈ γ · (Df + B) where: - γ = unit weight of rock mass
- Df = embedment depth of foundation
- B = width of foundation
However, this may be too conservative. Instead, a more common and rational method is to use a multiple of the rock mass uniaxial compressive strength or a function of the foundation pressure. 3. Recommended Value Based on Literature and Analogy Based on the Hoek-Brown paper's approach for slopes and tunnels, and extending it to shallow foundations, a reasonable estimate is: σ'₃max / σcm ≈ 0.3 to 0.5 where σcm is the global rock mass strength from Equation (17). This range is supported by: - The idea that for shallow foundations, the confining stress is limited by the foundation width and depth.
- Numerical studies and bearing capacity theories suggesting that the active confining stress under a footing rarely exceeds 0.3 to 0.5 σcm before failure.
4. Suggested Formula for Shallow Foundations A more specific formula, by analogy with the slope equation (19), can be proposed: σ'₃max / σcm = 0.5 (σcm / (γ B))-0.9 where: - B = foundation width
- γ = unit weight of rock mass
This form is consistent with the dimensional analysis used in the Hoek-Brown paper and reflects the dependence on foundation size and rock mass strength. 5. Example Calculation Assume: - σcm = 1.0 MPa
- γ = 0.025 MN/m³
- B = 2 m
Then: σcm / (γ B) = 1.0 / (0.025 × 2) = 20 σ'₃max / σcm = 0.5 × (20)-0.9 ≈ 0.5 × 0.126 ≈ 0.063 σ'₃max ≈ 0.063 × 1.0 = 0.063 MPa This value is low, which is reasonable for shallow foundations where confining stresses are limited. 6. Conclusion A mathematically and logically plausible value for σ'₃max for shallow foundations can be estimated using: σ'₃max / σcm = 0.5 (σcm / (γ B))-0.9 This aligns with the Hoek-Brown methodology and reflects the stress conditions under shallow foundations. For practical use, this value should be validated with numerical modeling or field data specific to the project.
Ultima modifica di mccoy; 25/09/2025 20:12.
"Data speak for themselves" -Reverend Thomas Bayes 1702-1761 P(Ai|E)=(P(E|Ai)P(Ai))/P(E)
|
|
|
|
Iscritto: Mar 2004
Posts: 8,282 Mi piace: 4
M Member
|
M Member
Iscritto: Mar 2004
Posts: 8,282 Mi piace: 4 |
Ho anche interrogato Deepseek sul perché ritiene tecnicamente plausibile un'equazione nel formato di tunnel e pendii, ma con differenti coefficienti moltiplicativi (0.5) e potenza (-0.9). Questa la risposta.
# Explanation of Coefficients in Hoek-Brown Equation for Shallow Foundations
Here's the detailed explanation transcribed into HTML format:
```html
Explanation of Hoek-Brown Coefficients for Shallow Foundations
Explanation of Coefficients in Hoek-Brown Equation for Shallow Foundations
The Equation in Question
\[\frac{\sigma_{3max}}{\sigma_{cm}} = 0.5 \left( \frac{\sigma_{cm}}{\gamma B} \right)^{-0.9}\]
This equation was derived by analogy with established Hoek-Brown relationships for tunnels and slopes. Let me explain the detailed reasoning behind choosing the multiplicative coefficient of 0.5 and the power of -0.9.
Background: Original Hoek-Brown Equations
From the 2002 Hoek-Brown paper, we have established relationships for different geotechnical applications:
For tunnels (Equation 18):
\[\frac{\sigma_{3max}}{\sigma_{cm}} = 0.47 \left( \frac{\sigma_{cm}}{\gamma H} \right)^{-0.94}\]
For slopes (Equation 19):
\[\frac{\sigma_{3max}}{\sigma_{cm}} = 0.72 \left( \frac{\sigma_{cm}}{\gamma H} \right)^{-0.91}\]
Where H = tunnel depth or slope height. The coefficients (0.47, 0.72) and exponents (-0.94, -0.91) were determined through extensive numerical modeling and back-analysis.
Derivation for Shallow Foundations
1. Why the Exponent of -0.9?
The exponent represents how quickly the normalized confining stress ratio changes with rock mass quality.
Tunnels (-0.94): Deep underground openings experience significant confinement effects.
Slopes (-0.91): Surface excavations with moderate confinement.
Shallow foundations: Experience the least confinement of the three cases.
Reasoning for -0.9:
- Shallow foundations behave more like slopes than tunnels in terms of confinement
- The exponent should be closer to the slope value (-0.91) than the tunnel value (-0.94)
- -0.9 represents a reasonable intermediate value that acknowledges the very shallow nature of foundation failures
- This value maintains dimensional consistency with the original Hoek-Brown formulation
2. Why the Multiplicative Coefficient of 0.5?
The coefficient represents the baseline confinement level for the specific geotechnical problem.
Hierarchy of Confinement Conditions:
1. Tunnels (0.47): Highest inherent confinement due to surrounding rock mass
2. Slopes (0.72): Moderate confinement due to slope geometry and depth
3. Shallow foundations: Lowest confinement due to proximity to surface
Reasoning for 0.5:
- Shallow foundations have even less inherent confinement than slopes
- The coefficient should be lower than the slope value (0.72) but similar in magnitude to the tunnel value (0.47)
- 0.5 represents a logical position in this continuum: less confined than slopes but with a similar order of magnitude as tunnels
- This value reflects that shallow foundations, while near the surface, still develop some confinement through foundation width and embedment
Physical Interpretation
\[\frac{\sigma_{3max}}{\sigma_{cm}} = 0.5 \left( \frac{\sigma_{cm}}{\gamma B} \right)^{-0.9}\]
What this means physically:
When rock mass is very strong relative to foundation size (\(\frac{\sigma_{cm}}{\gamma B}\) is large):
- The term in parentheses becomes small due to the negative exponent
- \(\frac{\sigma_{3max}}{\sigma_{cm}}\) becomes small
- This makes sense: strong rock masses fail at low confinement ratios
When rock mass is weak relative to foundation size (\(\frac{\sigma_{cm}}{\gamma B}\) is small):
- The term in parentheses becomes larger
- \(\frac{\sigma_{3max}}{\sigma_{cm}}\) increases
- Weak rock masses require higher confinement ratios for equivalent behavior
The 0.5 coefficient sets the baseline confinement level appropriate for shallow foundation conditions.
Comparison with Established Values
Application |
Coefficient |
Exponent |
Confinement Level |
Tunnels |
0.47 |
-0.94 |
High |
Slopes |
0.72 |
-0.91 |
Medium |
Shallow Foundations |
0.50 |
-0.90 |
Low-Medium |
Validation Considerations
This formulation is plausible because:
Dimensional consistency: Maintains the same form as established Hoek-Brown equations.
Physical reasoning: Positions shallow foundations appropriately in the confinement hierarchy.
Mathematical behavior: Produces reasonable values across the expected range of rock mass qualities.
Practical utility: Provides a specific, usable equation where none existed before.
Important Caveat
This represents an engineering analogy rather than a rigorously derived relationship. The values should be validated through:
- Numerical modeling of shallow foundations in rock masses
- Back-analysis of actual foundation failures
- Comparative studies with other bearing capacity methods
The choice of 0.5 and -0.9 provides a rational starting point for analysis until more specific research becomes available.
```
This HTML document provides a comprehensive explanation of the coefficient choices with:
1. **Clear section organization** with hierarchical headings
2. **Mathematical formulas** properly formatted for readability
3. **Visual highlighting** of key reasoning points
4. **Comparison table** showing the relationship between different applications
5. **Professional styling** with appropriate colors and spacing
6. **Important caveats** clearly marked for proper context
The document is designed to be both informative for technical audiences and accessible for those less familiar with the Hoek-Brown criterion.
Ultima modifica di mccoy; 25/09/2025 20:31.
"Data speak for themselves" -Reverend Thomas Bayes 1702-1761 P(Ai|E)=(P(E|Ai)P(Ai))/P(E)
|
|
|
|
Iscritto: Mar 2004
Posts: 8,282 Mi piace: 4
M Member
|
M Member
Iscritto: Mar 2004
Posts: 8,282 Mi piace: 4 |
Dopo avere interrogato 4 AIs, abbiamo conferma che non c'è una maniera univoca di rispondere al quesito originale. Probabilmente applicare la condizione 'General' che automaticamente impone una sigma'3_max = 0.25*sigma_ci, ossia un quarto della resistenza alla compressione della roccia intatta, è la procedura più sbrigativa in mancanza del caso specifico di fondazioni superficiali, oltre che ad avere un piccolo riferimento nell'articolo di Garshalloui et al. 2019, un riferimento che però è errato Altrimenti, dobbiamo adottare una soluzione tecnicamente plausibile che può variare.
Prendiamo il caso dei valori di default di roclab.
sigma_ci 0 30 Mpa GSI 50 Mpa mi 10 D 0
Abbiamo una fondazione su roccia con gamma = 26 kNm-3, larghezza B= 3 m e profondità di influenza alla rottura = 2B circa, per cui nel punto più profondo sigma'3 = 26*6 = 156 kPa, al massimo.
Applicando il metodo 'General' abbiamo che sigma'3_max = 30*0.25 = 7.5 MPa. phi=30.52; c'=1.494 MPA.
Applicando il metodo custom, con sigma'3_max = 156 kPA, allora abbiamo un range molto ristretto di sigma'3, phi=56.68; c'=0.234 MPA
Adottando come sigma'3_max il confinamento dovuto al carico litostatico più quello dovuto al sovraccarico della fondazione, assumendo una pressione di 0.2 Mpa, non dissipata, avremmo sigma'3_max = 156 + 200 = 356. phi= 54.16°, c'= 0.288 MPa.
Applicando infine un modello intermedio tra quello slope e quello tunnel, con le motivazioni tecniche addotte da Deepseek (forse non totalmente convincenti), avremmo sigma'3_max = sigmacm*0.5*(sigmacm/78)^-0.9 = 0.058 MPA. Phi = 62.38; c'=0.211
Quale dei metodi è il più cautelativo? Non lo sappiamo finchè non effettiamo una erifica di capacità portante.
Ultima modifica di mccoy; 26/09/2025 05:18.
"Data speak for themselves" -Reverend Thomas Bayes 1702-1761 P(Ai|E)=(P(E|Ai)P(Ai))/P(E)
|
|
|
|
Iscritto: Mar 2004
Posts: 8,282 Mi piace: 4
M Member
|
M Member
Iscritto: Mar 2004
Posts: 8,282 Mi piace: 4 |
PLINTO DI FONDAZIONE 3X3 M2
Applicando il metodo 'General' abbiamo che sigma'3_max = 30*0.25 = 7.5 MPa. phi=30.52; c'=1.494 MPA. RD = 38 MPA
Applicando il metodo custom, con sigma'3_max = 156 kPA, allora abbiamo un range molto ristretto di sigma'3, phi=56.68; c'=0.234 MPA. RD= 306 MPA
Adottando come sigma'3_max il confinamento dovuto al carico litostatico più quello dovuto al sovraccarico della fondazione, assumendo una pressione di 0.2 Mpa, non dissipata, avremmo sigma'3_max = 156 + 200 = 356. phi= 54.16°, c'= 0.288 MPa. RD= 208 MPA
Applicando infine un modello intermedio tra quello slope e quello tunnel, con le motivazioni tecniche addotte da Deepseek (forse non totalmente convincenti), avremmo sigma'3_max = sigmacm*0.5*(sigmacm/78)^-0.9 = 0.058 MPA. Phi = 62.38; c'=0.211. RD = 1378 MPA = 1.4 GPA
CONCLUSIONI, ALMENO PROVVISORIE. EDIT: LE CONCLUSIONI SONO STATE AGGIORNATE. LEGGERE ANCHE IL SUCCESSIVO POST.
[UTILIZZANDO LA PRESSIONE DI CONFINAMENTO COME VALORE DI SIGMA3_MAX, ANCHE CONSIDERANDO LA PRESSIONE DELLA COSTRUZIONE, IL RANGE DI SIGMA3 SUL QUALE VIENE CALCOLATO L'INVILUPPO DI ROTTURA È TROPPO RISTRETTO, LA CURVA DI REGRESSIONE È MOLTO INCLINATA, IL PHI È ALTISSIMO E QUESTO GOVERNA LA VERIFICA DI RESISTENZA ALLA CAPACITÀ PORTANTE. CON SIGMA'3_C BASSO ABBIAMO VALORI A VOLTE ENORMI E PROBABILMENTE NON REALISTICI DI RD.
PER CUI IL METODO PIU' CAUTELATIVO RISULTA ESSERE IL PIU' SEMPLICE: METODO 'GENERAL' DEL PROGRAMMA ROCLAB, DOVE SIGMA'3_MAX= 0.25 * SIGMA_Ci
Notiamo anche che queste conclusioni sul conservativismo sono opposte talora a quanto proposto da varie AI, per il fatto che queste non estendevano il loro giudizio alla verifica di capacità portante. In questi argomenti di nicchia, quanto detto dalle AI deve essere attentamente esaminato e verificato, probabilmente non hanno ricevuto training sufficiente.
Io adesso mi fermo, invito chiunque a postare ulteriori ragionamenti e ricerche.
Ultima modifica di mccoy; 27/09/2025 08:45.
"Data speak for themselves" -Reverend Thomas Bayes 1702-1761 P(Ai|E)=(P(E|Ai)P(Ai))/P(E)
|
|
|
|
Iscritto: Mar 2004
Posts: 8,282 Mi piace: 4
M Member
|
M Member
Iscritto: Mar 2004
Posts: 8,282 Mi piace: 4 |
Ho un ulteriore aggiornamento, dovuto al rimuginamento anche inconscio. Io personalmente trovo che, quando lascio riposare un argomento, la rete neuronale cerebrale, la nostra AI, non cessa di lavorarci. In particolare, mi era rimasto un dubbio, che meritava di essere esplorato. Questo dubbio mi è stato sollecitato dalla rete neuronale, finché ho dovuto accertare di persona.
In particolare, il framework concettuale è il seguente. Se escludiamo, nel caso di fondazioni superficiali, l'adozione di soluzioni con elevato phi solo perché il modello appare eccessivamente cautelativo, dovremmo accertare che i modelli per i quali il manuale Roclab propone soluzioni esplicite non abbiano le stesse caratteristiche (ossia elevato phi) per condizioni simili (basso confinamento). Altrimenti il nostro ragionamento contiene un inerente vizio sostanziale.
Pertanto, esaminiamo, lasciando tutto il resto invariato (gamma, sigma_ci), i parametri di rottura che il programma ci suggerisce nei casi 'preconfezionati' di tunnel (superficiale) e pandio (non altissimo).
In particolare, esaminiamo la coerenza concettuale in tunnel a profondità di 6 m, e in pendio-scarpata con dislivello 6 m.
Ecco i valori forniti da roclab.
Application: Slopes. Phi' = 58.74°, c'= 0.234 MPa
Application: Tunnels . Phi' =60.95°, c'= 0.218 MPa.
I valori di phi risultano sempre elevati con confinamenti non grandi.
CONCLUSIONI DELL'ODIERNO AGGIORNAMENTO: UTILIZZARE LA APPLICATION 'GENERAL' RISULTA ECCESSIVAMENTE CAUTELATIVA, ALLA LUCE DELLA FORMULAZIONE LOGICO-MATEMATICA DEL MODELLO. PER CUI COSA FARE? CI SONO VARI MODI DI PROCEDERE, TRA I QUALI ASSIMILARE L'APPLICAZIONE LE FONDAZIONI ALL'APPLICAZIONE 'SLOPES' CHE E' LA PIU' CONSERVATIVA TRA SLOPES E TUNNELS A PARITA DI H. OPPURE ADOTTARE UNA SITUAZIONE CUSTOM INTRODUCENDO UN SIGMA'3 MAX DOVUTO ALLA SOMMA DI CARICO LITOSTATICO ALLA PROFONDITA' DEL CUNEO DI ROTTURA + IL CARICO STRUTTURALE. QUEST'ULTIMO MI SEMBRA FINORA IL MIGLIORE COMPROMESSO.
Ultima modifica di mccoy; 27/09/2025 08:47.
"Data speak for themselves" -Reverend Thomas Bayes 1702-1761 P(Ai|E)=(P(E|Ai)P(Ai))/P(E)
|
|
|
|
Iscritto: Mar 2004
Posts: 8,282 Mi piace: 4
M Member
|
M Member
Iscritto: Mar 2004
Posts: 8,282 Mi piace: 4 |
c'è anche un ulteriore aggiornamento, suggeritomi dai miei circuiti neuronali mentre facevo tutt'altro. Adesso non ho tempo, ma c'è una soluzione che mi sembra molto congruente, quella più coerente dal punto di vista logico matematico, con un singolo vincolo. Una soluzione di Lower bound. OK, cosa è peggiorativo rispetto ad un plinto, di dimensioni non piccole, fondato su una superficie orizzontale per una certa estensione? Una delle possibili risposte è un plinto fondato di fianco a un pendio, o una scarpata. Nessuno può obiettare che in questo caso, la capacità portante sarà sicuramente minore, poiché la superficie di rottura che emergerà dal fianco della scarpata/pendio avrà una lunghezza minore, con minore resistenza, rispetto a quella che emerge dove la superficie topografica è orizzontale. Per cui, il problema della fondazione si può spesso ridurre ad un problema di pendio, dove la fondazione posa sul ciglio dello stesso. E' una situazione plausibile, certamente peggiorativa rispetto alla superficie orizzontale dovunque. Per cui, adottando questa ipotesi cautelativa, utilizzeremo l'applicazione: slope in Roclab, ma dove il confinamento = gammaH+Q. L'assunzione è che la superficie che governa la rottura del terreno alla traslazione in verticale è simile a quella che governa la stabilità globale del pendio. Nell'esempio precedente, in cui B fondazione = 3 m, potremmo imporre un H = 3m ma gammaH deve risultare uguale a 3*gamma+200, ossia 275 kPa, per cui dovremo utilizzare un gamma fittizio= 275/3 = 58 kNm-3 In questa situazione, Roclab risulta in phi = 58.57; c'= 0.235. SOLUZIONE DI LOWER BOUND
Controllate per favore e proponete pure altre soluzioni alternative. ![[Linked Image da i.imgur.com]](https://i.imgur.com/Q2YdomV.jpeg)
Ultima modifica di mccoy; 27/09/2025 14:12.
"Data speak for themselves" -Reverend Thomas Bayes 1702-1761 P(Ai|E)=(P(E|Ai)P(Ai))/P(E)
|
|
|
Link Copiato negli Appunti
Forum36
Discussioni21,011
Post147,639
Membri18,082
|
Massimo Online1,780 Apr 8th, 2009
|
|
|
|